• Re: Arms race

    From Martin Brown@'''newspam'''@nonad.co.uk to sci.electronics.design on Fri Nov 6 10:24:07 2026
    From Newsgroup: sci.electronics.design

    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and data. Signed drivers

    Not a Lisp fan then?

    People love to invent weird computer languages. There are over 3000
    and they keep coming.

    Real men code in binary using the front panel switches.

    I actually used a PDP-7 where you really did have to bootstrap it that
    way to read in a paper tape. It was a fun machine as it had a vector
    graphics display so was very good for asteroids, but cheated at chess.
    --
    Martin Brown

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From john larkin@jl@glen--canyon.com to sci.electronics.design on Fri Nov 6 03:22:01 2026
    From Newsgroup: sci.electronics.design

    On Fri, 6 Nov 2026 10:24:07 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and data. Signed drivers >>>
    Not a Lisp fan then?

    People love to invent weird computer languages. There are over 3000
    and they keep coming.

    Real men code in binary using the front panel switches.

    I actually used a PDP-7 where you really did have to bootstrap it that
    way to read in a paper tape. It was a fun machine as it had a vector >graphics display so was very good for asteroids, but cheated at chess.

    I had a PDP-8I with 4K of 12-bit core memory. One toggled in the RIM
    loader by hand, the first paper-tape reader bootstrap program. I had
    it memorized in octal.

    The machine cost $12K, teletype included, back when a cadillac cost
    maybe half that. It ran FOCAL, a cool interpretive language, and I
    simulated steamship propulsion systems on it.

    Our RP2040 chip costs 75 cents and computes maybe 10K times as fast.



    John Larkin
    Highland Tech Glen Canyon Design Center
    Lunatic Fringe Electronics
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From John R Walliker@jrwalliker@gmail.com to sci.electronics.design on Fri Nov 6 11:25:41 2026
    From Newsgroup: sci.electronics.design

    On 06/11/2026 10:24, Martin Brown wrote:
    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and data. Signed
    drivers

    Not a Lisp fan then?

    People love to invent weird computer languages. There are over 3000
    and they keep coming.

    Real men code in binary using the front panel switches.

    I actually used a PDP-7 where you really did have to bootstrap it that
    way to read in a paper tape. It was a fun machine as it had a vector graphics display so was very good for asteroids, but cheated at chess.


    I think that machine had a moon lander game as well.
    John
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Sat Nov 7 00:35:57 2026
    From Newsgroup: sci.electronics.design

    On 6/11/2026 9:24 pm, Martin Brown wrote:
    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and data. Signed
    drivers

    Not a Lisp fan then?

    People love to invent weird computer languages. There are over 3000
    and they keep coming.

    Real men code in binary using the front panel switches.

    I actually used a PDP-7 where you really did have to bootstrap it that
    way to read in a paper tape. It was a fun machine as it had a vector graphics display so was very good for asteroids, but cheated at chess.

    You had to hand-load the RIM loader, which could process the output of
    the paper tape reader, which you could then use to load the more
    efficient BIN loader on paper tape. Graduate student time was cheap.
    --
    Bill Sloman, Sydney

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Don Y@blockedofcourse@foo.invalid to sci.electronics.design on Fri Nov 6 07:31:10 2026
    From Newsgroup: sci.electronics.design

    On 11/6/2026 3:24 AM, Martin Brown wrote:
    Real men code in binary using the front panel switches.

    I used to have to write diagnostics for Novas and enter them with
    via the front panel, in octal. Hit RUN and *hope*.

    But, the diagnostics tended to be very short -- test the speech
    synthesizer, exercise the scanner motors, check core (!), etc.

    The "real" bootstrap was a 16 word ROM that loaded a digital
    cassette. Again, press IPL and *hope* (not many checks possible
    in a 16 word program!)

    OTOH, one the system was loaded, the next boot was literally
    instantaneous due to its persistence in core!

    Later models transitioned to semiconductor memory which, IMO, was a
    downgrade as the nonvolatile aspect was gone.

    I actually used a PDP-7 where you really did have to bootstrap it that way to
    read in a paper tape. It was a fun machine as it had a vector graphics display
    so was very good for asteroids, but cheated at chess.

    I used an Imlac PDS-1 for that. With a light-pen interface.

    <https://history-computer.com/inventions/pds-1-minicomputer/>

    Only commercial machine I've ever encountered with that sort of display
    and UI (though I designed a vector-graphic display system many years ago
    that had considerably more "ink" -- and wasn't constrained to "green"!)

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Simon Simple@nothanks@nottoday.co.uk to sci.electronics.design on Fri Nov 6 15:00:10 2026
    From Newsgroup: sci.electronics.design

    On 06/11/2026 11:25, John R Walliker wrote:
    On 06/11/2026 10:24, Martin Brown wrote:
    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and data. Signed
    drivers

    Not a Lisp fan then?

    People love to invent weird computer languages. There are over 3000
    and they keep coming.

    Real men code in binary using the front panel switches.

    I actually used a PDP-7 where you really did have to bootstrap it that
    way to read in a paper tape. It was a fun machine as it had a vector
    graphics display so was very good for asteroids, but cheated at chess.


    I think that machine had a moon lander game as well.
    John

    If that's the one I remember, you could successfully land if you
    plummeted till you were 1 metre from the surface then burned all your fuel.
    --
    SS

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Don Y@blockedofcourse@foo.invalid to sci.electronics.design on Fri Nov 6 08:27:30 2026
    From Newsgroup: sci.electronics.design

    On 11/6/2026 8:00 AM, Simon Simple wrote:
    On 06/11/2026 11:25, John R Walliker wrote:
    On 06/11/2026 10:24, Martin Brown wrote:
    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and data. Signed drivers >>>>>
    Not a Lisp fan then?

    People love to invent weird computer languages. There are over 3000
    and they keep coming.

    Real men code in binary using the front panel switches.

    I actually used a PDP-7 where you really did have to bootstrap it that way >>> to read in a paper tape. It was a fun machine as it had a vector graphics >>> display so was very good for asteroids, but cheated at chess.


    I think that machine had a moon lander game as well.
    John

    If that's the one I remember, you could successfully land if you plummeted till
    you were 1 metre from the surface then burned all your fuel.

    I wrote one for our Wang, multiuser "programmable calculator" (with nixie displays!) in high school. Inspired by a BASIC game I had seen years earlier.

    It was amusing as I created the various constraints without knowledge of whether or not you actually *could* land it!

    Aside from trivial strategies, I still don't know how I would go about
    ensuring it was solvable -- without padding the fuel budget or
    having other constraints on the rate of descent that would reduce the
    solution space.

    (obviously, with humans onboard, you'd want to constrain the forces
    to which they would be subjected and not just "hard brake" at the last
    instant)

    We also had an analog controller at school (as an interactive public
    display in the Aero&Astro building) that could be set to produce near
    chaotic behavior. It was a fun pasttime in the wee hours of the morning
    when walking across campus.

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Don Y@blockedofcourse@foo.invalid to sci.electronics.design on Fri Nov 6 08:40:50 2026
    From Newsgroup: sci.electronics.design

    On 11/6/2026 8:27 AM, Don Y wrote:
    On 11/6/2026 8:00 AM, Simon Simple wrote:

    If that's the one I remember, you could successfully land if you plummeted >> till you were 1 metre from the surface then burned all your fuel.

    I wrote one for our Wang, multiuser "programmable calculator" (with nixie displays!) in high school.-a Inspired by a BASIC game I had seen years earlier.

    It was amusing as I created the various constraints without knowledge of whether or not you actually *could* land it!
    <https://www.wangmuseum.nl/wp-content/uploads/2021/01/wang360.jpg>
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From john larkin@jl@htigct.com to sci.electronics.design on Fri Nov 6 10:54:08 2026
    From Newsgroup: sci.electronics.design

    On Fri, 6 Nov 2026 10:24:07 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and data. Signed drivers >>>
    Not a Lisp fan then?

    People love to invent weird computer languages. There are over 3000
    and they keep coming.

    Real men code in binary using the front panel switches.

    I actually used a PDP-7 where you really did have to bootstrap it that
    way to read in a paper tape. It was a fun machine as it had a vector >graphics display so was very good for asteroids, but cheated at chess.

    I did a lot of stuff in PDP-11 and 68K assembly languages. With a few
    macros, they are almost higher-level languages.

    Unlike some architectures that I might name.

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Sat Nov 7 15:38:13 2026
    From Newsgroup: sci.electronics.design

    On 7/11/2026 5:54 am, john larkin wrote:
    On Fri, 6 Nov 2026 10:24:07 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and data. Signed drivers >>>>
    Not a Lisp fan then?

    People love to invent weird computer languages. There are over 3000
    and they keep coming.

    Real men code in binary using the front panel switches.

    I actually used a PDP-7 where you really did have to bootstrap it that
    way to read in a paper tape. It was a fun machine as it had a vector
    graphics display so was very good for asteroids, but cheated at chess.

    I did a lot of stuff in PDP-11 and 68K assembly languages. With a few
    macros, they are almost higher-level languages.

    All high level languages get compiled into assembly language before than
    can be run on a computer.

    Being a compiler is tedious.

    Unlike some architectures that I might name.

    A language isn't an architecture. It has to be processed into something
    that can be run on hardware (which will have an architecture - it is constructed from components that have been connected by some kind of architect).
    --
    Bill Sloman, Sydney

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From john larkin@jl@glen--canyon.com to sci.electronics.design on Fri Nov 6 20:57:30 2026
    From Newsgroup: sci.electronics.design

    On Sat, 7 Nov 2026 15:38:13 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 7/11/2026 5:54 am, john larkin wrote:
    On Fri, 6 Nov 2026 10:24:07 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and data. Signed drivers >>>>>
    Not a Lisp fan then?

    People love to invent weird computer languages. There are over 3000
    and they keep coming.

    Real men code in binary using the front panel switches.

    I actually used a PDP-7 where you really did have to bootstrap it that
    way to read in a paper tape. It was a fun machine as it had a vector
    graphics display so was very good for asteroids, but cheated at chess.

    I did a lot of stuff in PDP-11 and 68K assembly languages. With a few
    macros, they are almost higher-level languages.

    All high level languages get compiled into assembly language before than
    can be run on a computer.

    Well, no.


    Being a compiler is tedious.

    Unlike some architectures that I might name.

    A language isn't an architecture. It has to be processed into something
    that can be run on hardware (which will have an architecture - it is >constructed from components that have been connected by some kind of >architect).

    If that post were compiled to assembly language, it would be a nop.


    John Larkin
    Highland Tech Glen Canyon Design Center
    Lunatic Fringe Electronics
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Sat Nov 7 17:11:49 2026
    From Newsgroup: sci.electronics.design

    On 7/11/2026 3:57 pm, john larkin wrote:
    On Sat, 7 Nov 2026 15:38:13 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 7/11/2026 5:54 am, john larkin wrote:
    On Fri, 6 Nov 2026 10:24:07 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and data. Signed drivers

    Not a Lisp fan then?

    People love to invent weird computer languages. There are over 3000
    and they keep coming.

    Real men code in binary using the front panel switches.

    I actually used a PDP-7 where you really did have to bootstrap it that >>>> way to read in a paper tape. It was a fun machine as it had a vector
    graphics display so was very good for asteroids, but cheated at chess.

    I did a lot of stuff in PDP-11 and 68K assembly languages. With a few
    macros, they are almost higher-level languages.

    All high level languages get compiled into assembly language before they
    can be run on a computer.

    Well, no.

    Counter-example?

    Being a compiler is tedious.

    Unlike some architectures that I might name.

    A language isn't an architecture. It has to be processed into something
    that can be run on hardware (which will have an architecture - it is
    constructed from components that have been connected by some kind of
    architect).

    If that post were compiled to assembly language, it would be a nop.

    As a compiler, you do seem to be sub-optimal, if not downright
    afunctional. Ignoring implications you don't like isn't an option for a compiler. There can be functions they can't parse - in your case don't
    want to parse - but that doesn't generate a nop.
    --
    Bill Sloman, Sydney

    --
    Bill Sloman Sydney

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From john larkin@jl@glen--canyon.com to sci.electronics.design on Sat Nov 7 07:39:08 2026
    From Newsgroup: sci.electronics.design

    On Sat, 7 Nov 2026 17:11:49 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 7/11/2026 3:57 pm, john larkin wrote:
    On Sat, 7 Nov 2026 15:38:13 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 7/11/2026 5:54 am, john larkin wrote:
    On Fri, 6 Nov 2026 10:24:07 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and data. Signed drivers

    Not a Lisp fan then?

    People love to invent weird computer languages. There are over 3000 >>>>>> and they keep coming.

    Real men code in binary using the front panel switches.

    I actually used a PDP-7 where you really did have to bootstrap it that >>>>> way to read in a paper tape. It was a fun machine as it had a vector >>>>> graphics display so was very good for asteroids, but cheated at chess. >>>>
    I did a lot of stuff in PDP-11 and 68K assembly languages. With a few
    macros, they are almost higher-level languages.

    All high level languages get compiled into assembly language before they >>> can be run on a computer.

    Well, no.

    Counter-example?

    Interpreters.

    Most Pythons

    Tokeizers/threaded code compilers

    The coolest case was probably the Burroughs systems


    "AI Overview

    Burroughs Corporation uniquely designed its computer systemsumost
    famously the Burroughs Large Systems starting with the B5000 in
    1961uto run the ALGOL programming language natively rather than
    assembler or Fortran."

    Their Algol compiler made machine code directly and was written in
    Algol! There was never any intermediate step.



    Being a compiler is tedious.

    Unlike some architectures that I might name.

    A language isn't an architecture. It has to be processed into something
    that can be run on hardware (which will have an architecture - it is
    constructed from components that have been connected by some kind of
    architect).

    If that post were compiled to assembly language, it would be a nop.

    As a compiler, you do seem to be sub-optimal, if not downright
    afunctional. Ignoring implications you don't like isn't an option for a >compiler. There can be functions they can't parse - in your case don't
    want to parse - but that doesn't generate a nop.

    I've written four RTOS's and three compilers. None of the compilers
    generated intermediate assembly.

    How about you?


    --
    Bill Sloman, Sydney

    John Larkin
    Highland Tech Glen Canyon Design Center
    Lunatic Fringe Electronics
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Sun Nov 8 15:44:12 2026
    From Newsgroup: sci.electronics.design

    On 8/11/2026 2:39 am, john larkin wrote:
    On Sat, 7 Nov 2026 17:11:49 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 7/11/2026 3:57 pm, john larkin wrote:
    On Sat, 7 Nov 2026 15:38:13 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 7/11/2026 5:54 am, john larkin wrote:
    On Fri, 6 Nov 2026 10:24:07 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and data. Signed drivers

    Not a Lisp fan then?

    People love to invent weird computer languages. There are over 3000 >>>>>>> and they keep coming.

    Real men code in binary using the front panel switches.

    I actually used a PDP-7 where you really did have to bootstrap it that >>>>>> way to read in a paper tape. It was a fun machine as it had a vector >>>>>> graphics display so was very good for asteroids, but cheated at chess. >>>>>
    I did a lot of stuff in PDP-11 and 68K assembly languages. With a few >>>>> macros, they are almost higher-level languages.

    All high level languages get compiled into assembly language before they >>>> can be run on a computer.

    Well, no.

    Counter-example?

    Interpreters.

    An interpreter is just a line by line compiler.

    Most Pythons

    So you don't understand what's going on.

    Tokeizers/threaded code compilers

    So you don't understand what's going on.

    The coolest case was probably the Burroughs systems


    "AI Overview

    Burroughs Corporation uniquely designed its computer systemsrComost
    famously the Burroughs Large Systems starting with the B5000 in
    1961rCoto run the ALGOL programming language natively rather than
    assembler or Fortran."

    Their Algol compiler made machine code directly and was written in
    Algol! There was never any intermediate step.

    Being a compiler is tedious.

    Borroughs had an Algol compiler, written in Algol. It was still a compiler.

    Compiled code has more lines/commands/op codes than the code from which
    is was compiled.

    Unlike some architectures that I might name.

    A language isn't an architecture. It has to be processed into something >>>> that can be run on hardware (which will have an architecture - it is
    constructed from components that have been connected by some kind of
    architect).

    If that post were compiled to assembly language, it would be a nop.

    As a compiler, you do seem to be sub-optimal, if not downright
    afunctional. Ignoring implications you don't like isn't an option for a
    compiler. There can be functions they can't parse - in your case don't
    want to parse - but that doesn't generate a nop.

    I've written four RTOS's and three compilers. None of the compilers
    generated intermediate assembly.

    Who said anything about "intermediate assembler"?

    The point about assembler is that it is human readable - groups of
    letters rather than numerical op codes presented as strings of octal or hexadecimal numbers. It is directly translatable into numerical opcodes
    by a simple look-up table.

    if you want to see what compiled code is actually doing it's usually
    easier to understand what you get by running the look-up process backwards.

    How about you?

    Not my sphere of interest, even though I seem to have a better grasp of
    what's involved than you do.
    --
    Bill Sloman, Sydney


    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From john larkin@jl@glen--canyon.com to sci.electronics.design on Sat Nov 7 21:15:37 2026
    From Newsgroup: sci.electronics.design

    On Sun, 8 Nov 2026 15:44:12 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 8/11/2026 2:39 am, john larkin wrote:
    On Sat, 7 Nov 2026 17:11:49 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 7/11/2026 3:57 pm, john larkin wrote:
    On Sat, 7 Nov 2026 15:38:13 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 7/11/2026 5:54 am, john larkin wrote:
    On Fri, 6 Nov 2026 10:24:07 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and data. Signed drivers

    Not a Lisp fan then?

    People love to invent weird computer languages. There are over 3000 >>>>>>>> and they keep coming.

    Real men code in binary using the front panel switches.

    I actually used a PDP-7 where you really did have to bootstrap it that >>>>>>> way to read in a paper tape. It was a fun machine as it had a vector >>>>>>> graphics display so was very good for asteroids, but cheated at chess. >>>>>>
    I did a lot of stuff in PDP-11 and 68K assembly languages. With a few >>>>>> macros, they are almost higher-level languages.

    All high level languages get compiled into assembly language before they >>>>> can be run on a computer.

    Well, no.

    Counter-example?

    Interpreters.

    An interpreter is just a line by line compiler.

    Most Pythons

    So you don't understand what's going on.

    Tokeizers/threaded code compilers

    So you don't understand what's going on.

    The coolest case was probably the Burroughs systems


    "AI Overview

    Burroughs Corporation uniquely designed its computer systemsumost
    famously the Burroughs Large Systems starting with the B5000 in
    1961uto run the ALGOL programming language natively rather than
    assembler or Fortran."

    Their Algol compiler made machine code directly and was written in
    Algol! There was never any intermediate step.

    Being a compiler is tedious.

    Borroughs had an Algol compiler, written in Algol. It was still a compiler.

    Compiled code has more lines/commands/op codes than the code from which
    is was compiled.

    Unlike some architectures that I might name.

    A language isn't an architecture. It has to be processed into something >>>>> that can be run on hardware (which will have an architecture - it is >>>>> constructed from components that have been connected by some kind of >>>>> architect).

    If that post were compiled to assembly language, it would be a nop.

    As a compiler, you do seem to be sub-optimal, if not downright
    afunctional. Ignoring implications you don't like isn't an option for a
    compiler. There can be functions they can't parse - in your case don't
    want to parse - but that doesn't generate a nop.

    I've written four RTOS's and three compilers. None of the compilers
    generated intermediate assembly.

    Who said anything about "intermediate assembler"?

    The point about assembler is that it is human readable - groups of
    letters rather than numerical op codes presented as strings of octal or >hexadecimal numbers. It is directly translatable into numerical opcodes
    by a simple look-up table.

    if you want to see what compiled code is actually doing it's usually
    easier to understand what you get by running the look-up process backwards.

    How about you?

    Not my sphere of interest, even though I seem to have a better grasp of >what's involved than you do.

    OK, what moron said

    "All high level languages get compiled into assembly language before
    than can be run on a computer."


    John Larkin
    Highland Tech Glen Canyon Design Center
    Lunatic Fringe Electronics
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Sun Nov 8 17:00:52 2026
    From Newsgroup: sci.electronics.design

    On 8/11/2026 4:15 pm, john larkin wrote:
    On Sun, 8 Nov 2026 15:44:12 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 8/11/2026 2:39 am, john larkin wrote:
    On Sat, 7 Nov 2026 17:11:49 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 7/11/2026 3:57 pm, john larkin wrote:
    On Sat, 7 Nov 2026 15:38:13 +1100, Bill Sloman <bill.sloman@ieee.org> >>>>> wrote:

    On 7/11/2026 5:54 am, john larkin wrote:
    On Fri, 6 Nov 2026 10:24:07 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and data. Signed drivers

    Not a Lisp fan then?

    People love to invent weird computer languages. There are over 3000 >>>>>>>>> and they keep coming.

    Real men code in binary using the front panel switches.

    I actually used a PDP-7 where you really did have to bootstrap it that >>>>>>>> way to read in a paper tape. It was a fun machine as it had a vector >>>>>>>> graphics display so was very good for asteroids, but cheated at chess. >>>>>>>
    I did a lot of stuff in PDP-11 and 68K assembly languages. With a few >>>>>>> macros, they are almost higher-level languages.

    All high level languages get compiled into assembly language before they >>>>>> can be run on a computer.

    Well, no.

    Counter-example?

    Interpreters.

    An interpreter is just a line by line compiler.

    Most Pythons

    So you don't understand what's going on.

    Tokeizers/threaded code compilers

    So you don't understand what's going on.

    The coolest case was probably the Burroughs systems


    "AI Overview

    Burroughs Corporation uniquely designed its computer systemsrComost
    famously the Burroughs Large Systems starting with the B5000 in
    1961rCoto run the ALGOL programming language natively rather than
    assembler or Fortran."

    Their Algol compiler made machine code directly and was written in
    Algol! There was never any intermediate step.

    Being a compiler is tedious.

    Borroughs had an Algol compiler, written in Algol. It was still a compiler. >>
    Compiled code has more lines/commands/op codes than the code from which
    is was compiled.

    Unlike some architectures that I might name.

    A language isn't an architecture. It has to be processed into something >>>>>> that can be run on hardware (which will have an architecture - it is >>>>>> constructed from components that have been connected by some kind of >>>>>> architect).

    If that post were compiled to assembly language, it would be a nop.

    As a compiler, you do seem to be sub-optimal, if not downright
    afunctional. Ignoring implications you don't like isn't an option for a >>>> compiler. There can be functions they can't parse - in your case don't >>>> want to parse - but that doesn't generate a nop.

    I've written four RTOS's and three compilers. None of the compilers
    generated intermediate assembly.

    Who said anything about "intermediate assembler"?

    The point about assembler is that it is human readable - groups of
    letters rather than numerical op codes presented as strings of octal or
    hexadecimal numbers. It is directly translatable into numerical opcodes
    by a simple look-up table.

    if you want to see what compiled code is actually doing it's usually
    easier to understand what you get by running the look-up process backwards. >>
    How about you?

    Not my sphere of interest, even though I seem to have a better grasp of
    what's involved than you do.

    OK, what moron said

    "All high level languages get compiled into assembly language before
    than can be run on a computer."

    It could well have been me. The translation between easily human
    readable alpha-numeric op-codes and the binary numbers that get loaded
    in the process-control register is a simple look-up, so "assembly
    language" is frequently understood to cover both. You clearly haven't
    reached that level of understanding, and it wouldn't be to your
    immediate advantage to admit it, even if you had finally got there.
    --
    Bill Sloman, Sydney

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From antispam@antispam@fricas.org (Waldek Hebisch) to sci.electronics.design on Sun Nov 8 19:27:09 2026
    From Newsgroup: sci.electronics.design

    Bill Sloman <bill.sloman@ieee.org> wrote:
    On 7/11/2026 3:57 pm, john larkin wrote:
    On Sat, 7 Nov 2026 15:38:13 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 7/11/2026 5:54 am, john larkin wrote:
    On Fri, 6 Nov 2026 10:24:07 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and data. Signed drivers

    Not a Lisp fan then?

    People love to invent weird computer languages. There are over 3000 >>>>>> and they keep coming.

    Real men code in binary using the front panel switches.

    I actually used a PDP-7 where you really did have to bootstrap it that >>>>> way to read in a paper tape. It was a fun machine as it had a vector >>>>> graphics display so was very good for asteroids, but cheated at chess. >>>>
    I did a lot of stuff in PDP-11 and 68K assembly languages. With a few
    macros, they are almost higher-level languages.

    All high level languages get compiled into assembly language before they >>> can be run on a computer.

    Well, no.

    Counter-example?

    Being a compiler is tedious.

    Unlike some architectures that I might name.

    A language isn't an architecture. It has to be processed into something
    that can be run on hardware (which will have an architecture - it is
    constructed from components that have been connected by some kind of
    architect).

    If that post were compiled to assembly language, it would be a nop.

    As a compiler, you do seem to be sub-optimal, if not downright
    afunctional. Ignoring implications you don't like isn't an option for a compiler. There can be functions they can't parse - in your case don't
    want to parse - but that doesn't generate a nop.

    You clearly go into area were you are not an expert. Google a
    bit about "undefined behaviour" in C and you will learn that
    when source code makes no sense, then C compier can do
    essentially anything, which in not so rare cases means that
    compiler simply deletes what programmer wrote. Explicit nop-s
    are rarely emited by compilers, but one can get 'return'
    (if compiler decides that garbage source is doing nothing
    useful, but it will eventually finish) or executing whatever
    happens to be after deleted code (if compiler decides that
    code should be never executed because it performed illegal
    operation).

    It is much nicer when you share things that you do know.
    --
    Waldek Hebisch
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Don Y@blockedofcourse@foo.invalid to sci.electronics.design on Sun Nov 8 14:00:12 2026
    From Newsgroup: sci.electronics.design

    On 11/8/2026 12:27 PM, Waldek Hebisch wrote:
    You clearly go into area were you are not an expert. Google a
    bit about "undefined behaviour" in C and you will learn that
    when source code makes no sense, then C compier can do
    essentially anything, which in not so rare cases means that
    compiler simply deletes what programmer wrote. Explicit nop-s

    The compiler doesn't have THAT much license. It can't decide to
    implement arbitrary operations just because you APPEAR to be
    relying on a behavior that is undefined.

    E.g., memcpy() is free to walk past the end of the address space
    if the source/destination address -- plus the length -- moves it
    to that point. Or, it could opt to be more reasonable and inhibit
    such action. Either solution is an acceptable behavior -- regardless
    of the behavior that you *hoped* would occur.

    But, it can't decide to divide every value stored in EVERY memory
    location by 3 -- unless the value stored there is a 7 in which case
    it replaces it with a 42. A compiler/library designer wouldn't
    deliberately come up with a bizarre action to perform when the required
    action isn't clear. He'll do what seems logical in a normal case,
    even if that isn't what the developer intended (developer broke the
    contract by trying to do something that wasn't defined; we don't
    punish him by killing his first-born...)
    are rarely emited by compilers, but one can get 'return'
    (if compiler decides that garbage source is doing nothing
    useful, but it will eventually finish) or executing whatever
    happens to be after deleted code (if compiler decides that
    code should be never executed because it performed illegal
    operation).

    What "illegal operation" -- beyond a SIGSEGV? The compiler
    doesn't prevent me from dividing by zero (though many are
    astute enough to notice that and WARN me). Or, from overwriting
    a table of interrupt vectors. Or, the software itself!

    It can *advise* against things that seem likely to be errors
    or result in anomalous behaviors. But, it's the developer who opts
    to override those warnings. No rule says you have to check malloc()'s
    return value. And, /in some cases/, you may be 100% confident that
    it is *NOT* NULL.

    [That's why folks who don't address every warning are shoddy
    developers -- make it explicitly clear to me and the compiler
    that you actually WANT this behavior!]
    It is much nicer when you share things that you do know.


    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Mon Nov 9 16:42:08 2026
    From Newsgroup: sci.electronics.design

    On 9/11/2026 6:27 am, Waldek Hebisch wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:
    On 7/11/2026 3:57 pm, john larkin wrote:
    On Sat, 7 Nov 2026 15:38:13 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 7/11/2026 5:54 am, john larkin wrote:
    On Fri, 6 Nov 2026 10:24:07 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and data. Signed drivers

    Not a Lisp fan then?

    People love to invent weird computer languages. There are over 3000 >>>>>>> and they keep coming.

    Real men code in binary using the front panel switches.

    I actually used a PDP-7 where you really did have to bootstrap it that >>>>>> way to read in a paper tape. It was a fun machine as it had a vector >>>>>> graphics display so was very good for asteroids, but cheated at chess. >>>>>
    I did a lot of stuff in PDP-11 and 68K assembly languages. With a few >>>>> macros, they are almost higher-level languages.

    All high level languages get compiled into assembly language before they >>>> can be run on a computer.

    Well, no.

    Counter-example?

    Being a compiler is tedious.

    Unlike some architectures that I might name.

    A language isn't an architecture. It has to be processed into something >>>> that can be run on hardware (which will have an architecture - it is
    constructed from components that have been connected by some kind of
    architect).

    If that post were compiled to assembly language, it would be a nop.

    As a compiler, you do seem to be sub-optimal, if not downright
    afunctional. Ignoring implications you don't like isn't an option for a
    compiler. There can be functions they can't parse - in your case don't
    want to parse - but that doesn't generate a nop.

    You clearly go into an area where you are not an expert.

    I'm not pretending to be an expert, just less of a clueless newbie in
    the area that John Larkin has shown himself to be.

    Google a bit about "undefined behaviour" in C.

    We weren't talking about "C" specifically - I do own a copy of Kernigan
    and Ritchie, but the last programing language I was fluent in was Fortran 4.

    A compiler is a computer program, and as Alan Turing proved around 1936,
    they are essentially unpredictable.

    https://www.cs.virginia.edu/~robins/Turing_Paper_1936.pdf

    It is much nicer when you share things that you do know.

    There are levels of knowledge, and software is an area where I am unsophisticated, but not entirely pig-ignorant.
    --
    Bill Sloman, Sydney

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From john larkin@jl@glen--canyon.com to sci.electronics.design on Mon Nov 9 07:12:58 2026
    From Newsgroup: sci.electronics.design

    On Mon, 9 Nov 2026 16:42:08 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 9/11/2026 6:27 am, Waldek Hebisch wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:
    On 7/11/2026 3:57 pm, john larkin wrote:
    On Sat, 7 Nov 2026 15:38:13 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 7/11/2026 5:54 am, john larkin wrote:
    On Fri, 6 Nov 2026 10:24:07 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and data. Signed drivers

    Not a Lisp fan then?

    People love to invent weird computer languages. There are over 3000 >>>>>>>> and they keep coming.

    Real men code in binary using the front panel switches.

    I actually used a PDP-7 where you really did have to bootstrap it that >>>>>>> way to read in a paper tape. It was a fun machine as it had a vector >>>>>>> graphics display so was very good for asteroids, but cheated at chess. >>>>>>
    I did a lot of stuff in PDP-11 and 68K assembly languages. With a few >>>>>> macros, they are almost higher-level languages.

    All high level languages get compiled into assembly language before they >>>>> can be run on a computer.

    Well, no.

    Counter-example?

    Being a compiler is tedious.

    Unlike some architectures that I might name.

    A language isn't an architecture. It has to be processed into something >>>>> that can be run on hardware (which will have an architecture - it is >>>>> constructed from components that have been connected by some kind of >>>>> architect).

    If that post were compiled to assembly language, it would be a nop.

    As a compiler, you do seem to be sub-optimal, if not downright
    afunctional. Ignoring implications you don't like isn't an option for a
    compiler. There can be functions they can't parse - in your case don't
    want to parse - but that doesn't generate a nop.

    You clearly go into an area where you are not an expert.

    I'm not pretending to be an expert, just less of a clueless newbie in
    the area that John Larkin has shown himself to be.

    I've written several compilers and several RTOS's and a bunch of
    system simulators and hundreds of application programs, a good chunk
    of a million lines of code. From PDP8 to ARM and one CPU that I
    designed myself.

    What's your definition of newbie?



    John Larkin
    Highland Tech Glen Canyon Design Center
    Lunatic Fringe Electronics
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Mon Nov 9 23:28:33 2026
    From Newsgroup: sci.electronics.design

    On 10/29/2026 8:59 PM, Don Y wrote:
    Management of these processes is the cause of most bad designs.
    The folks doing so are usually pretty ignorant of the efforts they
    are charged with managing!

    Please excuse me, I probably showed this example earlier this year in sci.electronics.design. A good example is

    https://neilmadden.blog/2022/04/19/psychic-signatures-in-java/

    where the /management/ was too lazy to hire a cryptographer for the pro-
    ject. It's not like you can accuse Oracle of not affording one.
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Mon Nov 9 23:31:54 2026
    From Newsgroup: sci.electronics.design

    On 10/31/2026 11:17 PM, Bill Sloman wrote:

    I don't make a habit of it, but there's no diplomatic way of getting the message across, when people are ignorant (which happens quite a lot), or stupid (which is rarer).

    Of course, you never see the same fault in yourself, you /fucking
    moron/!
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Mon Nov 9 23:40:11 2026
    From Newsgroup: sci.electronics.design

    On 10/30/2026 9:36 PM, Martin Brown wrote:
    I was always struck by the typical senior managers password being so
    weak that it was hardly worth having. Some of them even had it written
    on a postit and stuck to their screen! You couldn't make it up!

    By observation -- and this is something Bill Sloman never understands,
    he doesn't observe anything -- the fault lies with nobody of sufficient
    social status being around to /correct/ the managers.

    And since they are the ones hiring and firing the /security consultants/
    their own view is more important than whoever tries to teach them good
    password habits.

    Really, it's fine to write passwords down; just do it in a book you keep
    in your pocket, and don't forget it by the computer [terminal].

    And once you start writing the passwords down, you don't need to be
    afraid of the cryptographically secure random password generators.

    2FA helps if the password gets stolen somehow; it's usually not lifted
    from the book. It's still annoying.
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Mon Nov 9 23:47:29 2026
    From Newsgroup: sci.electronics.design

    On 10/30/2026 10:29 PM, Don Y wrote:
    What's amusing is all of these ideas are 50+ years old.-a Yet, people keep convincing themselves that they "need" more freedom... "constraints are
    bad"
    (as is the code they develop!)

    What is not amusing is that nobody thought to put computer security
    education in the elementary schools fifty years ago.

    And because the didn't do it then, they won't do it now -- is by obser-
    vation, their thinking. I don't know what goes on in the heads of peo-
    ple who decide the elementary curriculum. Usefulness in life is not
    their agenda.

    Anyway, a /large/ part of the problem is that now that everyone uses
    computers, the computer security practices people should have been
    taught in elementary school but weren't, don't work. Because they don't
    know them, and they don't know how real life organized criminals work.

    Err, for an introduction to that, you could use the book /Undercover
    Cop/ by Mike Russel and Patrick W. Picciarelli. It's an autobiography
    of sorts, and therefore describes the real world, and not the fantasy
    world in movies. You can then pair that knowledge with Nicole Perl-
    roth's book /This Is How They Tell Me the World Ends/, for more real
    world stories of computer security shenanigans.

    Because if you don't know who the enemy is, and how he operates, how are
    you going to secure your computer? With a ritual and magic smoke?
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Mon Nov 9 23:56:23 2026
    From Newsgroup: sci.electronics.design

    On 10/30/2026 11:47 PM, Don Y wrote:
    On 10/30/2026 8:36 AM, JM wrote:
    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and data. Signed drivers

    Not a Lisp fan then?

    You don't need *strict* separation (e.g., like a Harvard architecture)
    to benefit from this practice.-a If you can locate data/code on page boundaries and in page-sized units, you can commingle them freely.

    But, the idea of treating memory as an amorphous gel that can be
    freely molded is not going to work.

    Nowadays, you actually need to put the different types of memories
    in different *chips* if you truly want to maximize your protection
    (you can gain some advantages with guard bands between regions but
    the hardware has proven to easily exploitable in its efforts to
    try to eke out more performance)

    I don't know any hardware manufacturer doing this -- well, I don't work
    with military hardware so they could be doing it -- and now that Sam
    Altman is gobbling up all the DRAM chips, I doubt we'll see this in end
    user computers and phones any time soon.

    So, is this a loss due to the 2026 DRAM shortage? Only speculative his-
    tory in the future will tell us.
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Don Y@blockedofcourse@foo.invalid to sci.electronics.design on Mon Nov 9 10:42:28 2026
    From Newsgroup: sci.electronics.design

    On 11/9/2026 8:56 AM, Johann 'Myrkraverk' Oskarsson wrote:
    On 10/30/2026 11:47 PM, Don Y wrote:
    On 10/30/2026 8:36 AM, JM wrote:
    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and data. Signed drivers >>>
    Not a Lisp fan then?

    You don't need *strict* separation (e.g., like a Harvard architecture)
    to benefit from this practice.-a If you can locate data/code on page
    boundaries and in page-sized units, you can commingle them freely.

    But, the idea of treating memory as an amorphous gel that can be
    freely molded is not going to work.

    Nowadays, you actually need to put the different types of memories
    in different *chips* if you truly want to maximize your protection
    (you can gain some advantages with guard bands between regions but
    the hardware has proven to easily exploitable in its efforts to
    try to eke out more performance)

    I don't know any hardware manufacturer doing this -- well, I don't work
    with military hardware so they could be doing it -- and now that Sam
    Altman is gobbling up all the DRAM chips, I doubt we'll see this in end
    user computers and phones any time soon.

    You can often employ this technique in many of the larger SoCs.
    sadly, the resources are often allocated by someone who doesn't
    understand the VALUE of specific allocations.

    E.g., if you put kernel memory in SRAM then "other memory"
    being in LPDRAM has less potential -- via direct or side-channel
    attack -- of corrupting it.

    But, you have to be aware of the problem and understand how the
    hardware works before you can make smart decisions about it.
    Which, of course, means you have to have a firm hardware model
    before you can sort out the software.

    The hardware people are similarly clueless as they are usually not aware
    of how the hardware will be used (and abused). They often treat
    hardware as checkoff items: 2 serial ports; check. 3GB RAM; check.
    1MB boot ROM; check.

    One of the colossal FAILS IME was a hardware person putting a one-bit
    latch to drive the input of a CVSD. So, the software had to sit in a
    tight loop shifting bytes into the latch and clocking the CVSD.

    And, of course, every path through the code had to have identical time
    lest the clock frequency beat at the byte-rate.

    Really? You couldn't put a byte-wide (parallel load, serial out)
    SHIFT REGISTER there and have it self clocked so all the CPU has to
    do is push BYTES at it every 8 bit times??? You've traded a few pennies
    for the entire CPU (because it can't do ANYTHING ELSE while caught
    in this idiot loop).

    Perhaps we should let hardware people write code and publicly
    critique their failures? Or, let software people design the
    hardware and let the hardware folks write the code??

    So, is this a loss due to the 2026 DRAM shortage?-a Only speculative his- tory in the future will tell us.

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Don Y@blockedofcourse@foo.invalid to sci.electronics.design on Mon Nov 9 11:11:07 2026
    From Newsgroup: sci.electronics.design

    On 11/9/2026 8:47 AM, Johann 'Myrkraverk' Oskarsson wrote:
    On 10/30/2026 10:29 PM, Don Y wrote:
    What's amusing is all of these ideas are 50+ years old.-a Yet, people keep >> convincing themselves that they "need" more freedom... "constraints are bad" >> (as is the code they develop!)

    What is not amusing is that nobody thought to put computer security
    education in the elementary schools fifty years ago.

    Expecting society to embrace "computers" that long ago is probably a stretch.

    I started writing code about 55 years ago. "Security" was:
    - THE machine was in its own room
    - you kept control of your card deck

    The Computer was most often a tool for business; we (Engineering) had
    to beg for time from Accounting to run any of our algorithms on "their" machine.

    And because the didn't do it then, they won't do it now -- is by obser- vation, their thinking.-a I don't know what goes on in the heads of peo-
    ple who decide the elementary curriculum.-a Usefulness in life is not
    their agenda.

    I have exactly one data point in that regard.

    A neighbor is responsible for curriculum, locally (I believe there is
    a committee of sorts). I once mentioned colorblindness, in passing,
    and cited the "1 in ~15" men are colorblind statistic.

    She *refused* to believe me -- after all, that means at least one
    "boy" in every class (on average) can't tell the "red apples" from
    the "green apples" (red-green colorblindness being the most common;
    think about that when you have a GREEN "start" button and a RED
    "stop" button on a device!) in the sample problems they are given.

    Her ignorance left me speechless -- you're making decisions about
    courseware and textbooks and you don't even know what your
    "customers" need???

    She was equally indignant and brought it up on some online forum.

    Only to be told of HER error. In a very public way.

    Anyway, a /large/ part of the problem is that now that everyone uses computers, the computer security practices people should have been
    taught in elementary school but weren't, don't work.-a Because they don't know them, and they don't know how real life organized criminals work.

    But that's part of human nature. People aren't very smart nor sophisticated. Confidence trumps their fear of showing their ignorance so if you present something as fact, they are more likely to agree than to have to think
    for themselves (esp if that thinking will be visible in such a discussion)

    I received a fishing email from a friend many months ago:
    "I need to borrow $400 for my niece's birthday gift. I'm not
    at my phone now. Can you buy me a gift card and read me the
    code number off it?"
    Gee, you MAY have a niece (I'm not familiar with your extended family).
    You're NEVER "away from your phone".
    You'd never make such an "ask" in an email (instead of by phone or in person) You only live a couple of miles from here yet can't be inconvenienced to
    drive over FOR THE CASH.
    Your name is misspelled in the email headed.

    "Sure! I'll head over right now!"
    "No. Just read me the numbers."
    "Ah, but I'm just down the road and its been a while since we've visited..."

    Meanwhile, I'm dialing his phone to tell him his email account has been
    hacked.

    And, of course, all the cretins desperate for attention that they share
    every aspect of their lives on social media.

    "I don't understand. How did he know I had a grandson in college?"

    How did he know my birthdate? Mother's maiden name? etc.

    Err, for an introduction to that, you could use the book /Undercover
    Cop/ by Mike Russel and Patrick W. Picciarelli.-a It's an autobiography
    of sorts, and therefore describes the real world, and not the fantasy
    world in movies.-a You can then pair that knowledge with Nicole Perl-
    roth's book /This Is How They Tell Me the World Ends/, for more real
    world stories of computer security shenanigans.

    Because if you don't know who the enemy is, and how he operates, how are
    you going to secure your computer?-a With a ritual and magic smoke?

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Don Y@blockedofcourse@foo.invalid to sci.electronics.design on Mon Nov 9 11:23:54 2026
    From Newsgroup: sci.electronics.design

    On 11/9/2026 8:28 AM, Johann 'Myrkraverk' Oskarsson wrote:
    On 10/29/2026 8:59 PM, Don Y wrote:
    Management of these processes is the cause of most bad designs.
    The folks doing so are usually pretty ignorant of the efforts they
    are charged with managing!

    Please excuse me, I probably showed this example earlier this year in sci.electronics.design.-a A good example is

    -a https://neilmadden.blog/2022/04/19/psychic-signatures-in-java/

    where the /management/ was too lazy to hire a cryptographer for the pro- ject.-a It's not like you can accuse Oracle of not affording one.

    This (and the opposite: hiring someone and then ignoring their "expert recommendations") happens surprisingly often.

    The people making the decisions aren't aware of how much they are unaware. Dunning-Kruger. Particularly dangerous when The Ignorant are in charge.

    Nowhere is this more evident than in software as ANYONE can write code
    and, true to D-K, they all *think* they know far more than they actually
    do. "I've written..." yeah, but were any of those really more than trivial exercises? I.e., if it was the sort of problem we'd get in a homework assignment, you know it has a complexity on the order of 8 hours or less
    (cuz the homework budget for a course was 8 hours)

    Another favorite is "real-time" -- particularly "hard" vs. "soft".
    When most folks think it means "real fast" (it doesn't). Or, that
    "hard" is "more difficult" than "soft" (the opposite is true -- in
    spades -- which is why so many newbies resort to HRT solutions
    to problems... they can't think beyond FORCING a hard requirement
    when none is required... even in nuclear defenses!)

    But, so much of this is ingrained that challenging it threatens
    most folks' notion of their own competence so they ignore learning
    about those things where their ignorance is so profound.

    [You have to learn to work AROUND such people lest they compromise
    any good work]
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Tue Nov 10 02:25:00 2026
    From Newsgroup: sci.electronics.design

    On 11/10/2026 1:42 AM, Don Y wrote:
    On 11/9/2026 8:56 AM, Johann 'Myrkraverk' Oskarsson wrote:
    On 10/30/2026 11:47 PM, Don Y wrote:
    On 10/30/2026 8:36 AM, JM wrote:
    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and data. Signed
    drivers

    Not a Lisp fan then?

    You don't need *strict* separation (e.g., like a Harvard architecture)
    to benefit from this practice.-a If you can locate data/code on page
    boundaries and in page-sized units, you can commingle them freely.

    But, the idea of treating memory as an amorphous gel that can be
    freely molded is not going to work.

    Nowadays, you actually need to put the different types of memories
    in different *chips* if you truly want to maximize your protection
    (you can gain some advantages with guard bands between regions but
    the hardware has proven to easily exploitable in its efforts to
    try to eke out more performance)

    I don't know any hardware manufacturer doing this -- well, I don't work
    with military hardware so they could be doing it -- and now that Sam
    Altman is gobbling up all the DRAM chips, I doubt we'll see this in end
    user computers and phones any time soon.

    You can often employ this technique in many of the larger SoCs.
    sadly, the resources are often allocated by someone who doesn't
    understand the VALUE of specific allocations.

    E.g., if you put kernel memory in SRAM then "other memory"
    being in LPDRAM has less potential -- via direct or side-channel
    attack -- of corrupting it.

    But, you have to be aware of the problem and understand how the
    hardware works before you can make smart decisions about it.
    Which, of course, means you have to have a firm hardware model
    before you can sort out the software.

    The hardware people are similarly clueless as they are usually not aware
    of how the hardware will be used (and abused).-a They often treat
    hardware as checkoff items:-a 2 serial ports; check.-a-a 3GB RAM; check.
    1MB boot ROM; check.

    One of the colossal FAILS IME was a hardware person putting a one-bit
    latch to drive the input of a CVSD.-a So, the software had to sit in a
    tight loop shifting bytes into the latch and clocking the CVSD.

    And, of course, every path through the code had to have identical time
    lest the clock frequency beat at the byte-rate.

    Really?-a You couldn't put a byte-wide (parallel load, serial out)
    SHIFT REGISTER there and have it self clocked so all the CPU has to
    do is push BYTES at it every 8 bit times???-a You've traded a few pennies
    for the entire CPU (because it can't do ANYTHING ELSE while caught
    in this idiot loop).

    Perhaps we should let hardware people write code and publicly
    critique their failures?-a Or, let software people design the
    hardware and let the hardware folks write the code??

    So, is this a loss due to the 2026 DRAM shortage?-a Only speculative his-
    tory in the future will tell us.


    Or have more generalists around. I'm one -- no matter what Bill Sloman
    tries to say -- but with only a passing experience of hardware design.

    Just enough to understand the basics.

    If you dig into old Linus-rants on the LKML, you'll find the GCC-devs of
    the time didn't care how the compiler was used for anything other than
    user space. That's not a healthy environment for kernel devs.

    So we should let compilers devs do operating systems, and operating sys-
    tem people do compilers, and both groups do hardware.

    At minimum.


    Best wishes, and happy hardware design!
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Don Y@blockedofcourse@foo.invalid to sci.electronics.design on Mon Nov 9 12:39:27 2026
    From Newsgroup: sci.electronics.design

    On 11/9/2026 11:25 AM, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 1:42 AM, Don Y wrote:
    On 11/9/2026 8:56 AM, Johann 'Myrkraverk' Oskarsson wrote:
    On 10/30/2026 11:47 PM, Don Y wrote:
    On 10/30/2026 8:36 AM, JM wrote:
    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and data. Signed drivers >>>>>
    Not a Lisp fan then?

    You don't need *strict* separation (e.g., like a Harvard architecture) >>>> to benefit from this practice.-a If you can locate data/code on page
    boundaries and in page-sized units, you can commingle them freely.

    But, the idea of treating memory as an amorphous gel that can be
    freely molded is not going to work.

    Nowadays, you actually need to put the different types of memories
    in different *chips* if you truly want to maximize your protection
    (you can gain some advantages with guard bands between regions but
    the hardware has proven to easily exploitable in its efforts to
    try to eke out more performance)

    I don't know any hardware manufacturer doing this -- well, I don't work
    with military hardware so they could be doing it -- and now that Sam
    Altman is gobbling up all the DRAM chips, I doubt we'll see this in end
    user computers and phones any time soon.

    You can often employ this technique in many of the larger SoCs.
    sadly, the resources are often allocated by someone who doesn't
    understand the VALUE of specific allocations.

    E.g., if you put kernel memory in SRAM then "other memory"
    being in LPDRAM has less potential -- via direct or side-channel
    attack -- of corrupting it.

    But, you have to be aware of the problem and understand how the
    hardware works before you can make smart decisions about it.
    Which, of course, means you have to have a firm hardware model
    before you can sort out the software.

    The hardware people are similarly clueless as they are usually not aware
    of how the hardware will be used (and abused).-a They often treat
    hardware as checkoff items:-a 2 serial ports; check.-a-a 3GB RAM; check.
    1MB boot ROM; check.

    One of the colossal FAILS IME was a hardware person putting a one-bit
    latch to drive the input of a CVSD.-a So, the software had to sit in a
    tight loop shifting bytes into the latch and clocking the CVSD.

    And, of course, every path through the code had to have identical time
    lest the clock frequency beat at the byte-rate.

    Really?-a You couldn't put a byte-wide (parallel load, serial out)
    SHIFT REGISTER there and have it self clocked so all the CPU has to
    do is push BYTES at it every 8 bit times???-a You've traded a few pennies
    for the entire CPU (because it can't do ANYTHING ELSE while caught
    in this idiot loop).

    Perhaps we should let hardware people write code and publicly
    critique their failures?-a Or, let software people design the
    hardware and let the hardware folks write the code??

    So, is this a loss due to the 2026 DRAM shortage?-a Only speculative his- >>> tory in the future will tell us.

    Or have more generalists around.-a I'm one -- no matter what Bill Sloman tries to say -- but with only a passing experience of hardware design.

    Just enough to understand the basics.

    Just enough to understand WHO/WHAT you REALLY need to tackle a problem.

    If you dig into old Linus-rants on the LKML, you'll find the GCC-devs of
    the time didn't care how the compiler was used for anything other than
    user space.-a That's not a healthy environment for kernel devs.

    So we should let compilers devs do operating systems, and operating sys-
    tem people do compilers, and both groups do hardware.

    At minimum.
    "Impedance mismatches" in skillsets are common. Not just in design/development but, also, in userworld.

    If you offer a product that intimidates -- or *suggests* usage other than actual -- your customer expresses their frustration (cuz they don't want to admit "it's too complicated") by bad-mouthing/discarding/returning your product. Or, at the very least, grumbling about its use.

    We have a washing machine with "a big (stopless) knob" used to select the
    type of wash cycle. It has a textured grip and has a cogged feel when
    rotated.

    But, the cogs don't correspond to the points at which the "current selection" advances to the next choice. So, it always feels like something is broken... some linkage between the knob and the mechanism.

    And, the selection is indicated by an LED. So, not only does the "feel"
    FAIL to provide you with feedback as to how much you've turned the knob,
    but you also have to LOOK at the knob to see where it is, presently.

    Ditto the buttons that toggle the water temperature (wash/rinse), spin,
    soil level, "alert", etc. settings require your attention to VERIFY
    their current settings -- as turning the big knob will alter THEIR
    settings (possibly even invalidating some of the available choices).

    I.e., the message is: you must set the controls in the order some dweeb
    in our engineering department decided was appropriate -- as HE has seen
    fit to impose his constraints on your choices, after the fact.

    Contrast this with an old appliance where you turned a MECHANICAL (timer)
    to a desired setting and pressed "start" (often part of the same mechanism
    so accomplished with one had).

    This because the designer considered themselves to be the expert instead of
    the intended user.

    Or, worse, because the designer had total disregard for the user.

    If you don't know what to do OR don't understand the issues, then find
    someone who DOES!
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From antispam@antispam@fricas.org (Waldek Hebisch) to sci.electronics.design on Mon Nov 9 21:14:53 2026
    From Newsgroup: sci.electronics.design

    Don Y <blockedofcourse@foo.invalid> wrote:
    On 11/8/2026 12:27 PM, Waldek Hebisch wrote:
    You clearly go into area were you are not an expert. Google a
    bit about "undefined behaviour" in C and you will learn that
    when source code makes no sense, then C compier can do
    essentially anything, which in not so rare cases means that
    compiler simply deletes what programmer wrote. Explicit nop-s

    The compiler doesn't have THAT much license. It can't decide to
    implement arbitrary operations just because you APPEAR to be
    relying on a behavior that is undefined.

    Compiler just uses rules, so writing about appearence makes no
    sense. When optimizing compier performs a lot of transformations.
    Validity of transformations may depend on conditions. When
    checking conditions compiler assumes that program does not
    contain undefined behaviour.

    And yes, C standard gives such freedom to authors of C
    compilers:

    : undefined behavior
    : behavior, upon use of a nonportable or erroneous program construct
    : or of erroneous data, for which this document imposes no requirements

    Do I need to explain to you what "no requirements" mean?

    E.g., memcpy() is free to walk past the end of the address space
    if the source/destination address -- plus the length -- moves it
    to that point. Or, it could opt to be more reasonable and inhibit
    such action. Either solution is an acceptable behavior -- regardless
    of the behavior that you *hoped* would occur.

    But, it can't decide to divide every value stored in EVERY memory
    location by 3 -- unless the value stored there is a 7 in which case
    it replaces it with a 42. A compiler/library designer wouldn't
    deliberately come up with a bizarre action to perform when the required action isn't clear. He'll do what seems logical in a normal case,
    even if that isn't what the developer intended (developer broke the
    contract by trying to do something that wasn't defined; we don't
    punish him by killing his first-born...)

    What you write is typical fallacy. Of course compiler writers
    will not deliberatly generate malicious code. But once your
    part of contract is broken there is no reasonable way to
    limit possible damage. If compiler recognizes one of common
    errors it can warn and generate code that say signals
    error at runtime. But it is pretty commmon situation that
    compiler can not check if code is free of undefined behaviour.
    In such cases compiler transforms the code assuming that
    code is free of undefined behaviour. Elementary logic
    says that from false assumption you can prove anything.
    So, when there is undefined behaviour, there is no limit
    on wrong reasonings that compiler can make and consequently
    no limit of possible transformations. Of course, any
    given compiler contains only finitely many transformations
    and logical reasonings in the compiler are rather simple.
    But one of simplest transformations is "delete code that
    is never executed". And related "if condition of 'if'
    statement is always true, then 'else' branch is never
    executed'. So, one you have undefined behaviour compiler
    may easily delete large part of your program. In effect
    if control passes to a function which should be not
    called execution may "fall trough" to following code.
    If that code is doing something weird, you will see this
    as effect of your program.

    This was already understood in seventies: either you define
    what happens or you get potentially arbitrary conseqences.
    Defining things may get you one step ahead, but in many
    cases defined behaviour is wrong. PL/1 tried hard to define
    effects of various operations, users found out that many
    cases operations were well defined, but did different thing
    than what was needed and programs behaved in unexpected ways.
    Catching errors at runtime could help, but in the past
    it was deemed to costly. And currently, in C (which is
    widely used) catching errors at runtime is much more
    costly than in other languages like Pascal or Ada
    (or modern Fortran if you prefer that). C++ is better,
    as long as you avoid constructs inherited from C.
    Anyway, except for cases like Rust and Ada current system
    programming languages basically force programmers to
    use potentially unsafe constructs with possible
    consequences equvalent to undefined behaviour in C.

    are rarely emited by compilers, but one can get 'return'
    (if compiler decides that garbage source is doing nothing
    useful, but it will eventually finish) or executing whatever
    happens to be after deleted code (if compiler decides that
    code should be never executed because it performed illegal
    operation).

    What "illegal operation" -- beyond a SIGSEGV? The compiler
    doesn't prevent me from dividing by zero (though many are
    astute enough to notice that and WARN me). Or, from overwriting
    a table of interrupt vectors. Or, the software itself!

    Actually, at '-O2' version of 'gcc' that I use replaces explicit
    division by 0 by the 'ud2' instruction. But that missed the point,
    when you derefernce a pointer gcc assumes that it is nonzero,
    when you divide that divisor is nonzero. Importantly, gcc
    assumes that operations on integers do not overflow. Typically
    compiler can not prove at compile that if anything above
    happens, so in first approximation it just generates corresponding
    machine instrictions. But implied conditions may lead to
    conseqences, like deleting code which is unreachable by
    valid execution.

    If you want list of "illegal operations", then appendix in C
    standard contains rather long list of things which are undefined
    behaviour, trying any of them is illegal from standard point
    of view. gcc uses only some of them for optimization, like
    the few I mention above and the type based aliasing rules.
    But new release/different compiler may take advantage of
    a different undefined behaviour. So, if you use a single
    compiler look at what promises it makes, if you want
    wide portability then you should avoid all undefiend
    behaviour.

    It can *advise* against things that seem likely to be errors
    or result in anomalous behaviors. But, it's the developer who opts
    to override those warnings. No rule says you have to check malloc()'s
    return value. And, /in some cases/, you may be 100% confident that
    it is *NOT* NULL.

    [That's why folks who don't address every warning are shoddy
    developers -- make it explicitly clear to me and the compiler
    that you actually WANT this behavior!]
    It is much nicer when you share things that you do know.


    --
    Waldek Hebisch
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From =?UTF-8?B?Q8OzaWzDrW4=?= =?UTF-8?B?IE5pb2Nsw6Fzw61u?= =?UTF-8?B?IEdsb3N0w6lpcg==?=@thanks-to@Taf.com to sci.electronics.design on Mon Nov 9 22:36:39 2026
    From Newsgroup: sci.electronics.design

    Books and phones can become stolen.
    (S. HTTP://Gloucester.Insomnia247.NL/ fuer Kontaktdaten!)
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From =?UTF-8?B?Q8OzaWzDrW4=?= =?UTF-8?B?IE5pb2Nsw6Fzw61u?= =?UTF-8?B?IEdsb3N0w6lpcg==?=@thanks-to@Taf.com to sci.electronics.design on Mon Nov 9 23:01:45 2026
    From Newsgroup: sci.electronics.design

    Johann 'Myrkraverk' Oskarsson <johann@Myrkraverk.invalid> skrev: |----------------------------------------------------------------------|
    |"[. . .] |
    |What is not amusing is that nobody thought to put computer security | |education in the elementary schools fifty years ago." | |----------------------------------------------------------------------|

    It is not plausible that a computer-security education in the
    elementary schools fifty years ago (in 1976) would be useful. Almost
    no child in 1976 used to have an Apple I, and an Apple I does not
    demand a password. Cf.
    HTTP://WWW.oldComputers.net

    Johann 'Myrkraverk' Oskarsson <johann@Myrkraverk.invalid> skrev: |----------------------------------------------------------------------|
    |"[. . .] |
    |[. . .] I don't know what goes on in the heads of peo- |
    |ple who decide the elementary curriculum. Usefulness in life is not |
    |their agenda. |
    | |
    |[. . .]" | |----------------------------------------------------------------------|

    Of course it is! Elementary schools teach useful mathematics and
    useful words.

    Johann 'Myrkraverk' Oskarsson <johann@Myrkraverk.invalid> skrev: |----------------------------------------------------------------------|
    |"Err, for an introduction to that, you could use the book /Undercover |
    |Cop/ by Mike Russel and Patrick W. Picciarelli. It's an autobiography|
    |of sorts, and therefore describes the real world, and not the fantasy |
    |world in movies. [. . .] |
    |[. . .] |
    | |
    |[. . .]" | |----------------------------------------------------------------------|

    Thanks for this suggestion. How did Mike Russell survive "being
    ambushed and shot in the back of the head"?

    Is the next quotation true? "his police handlers disregarded his
    wishes and his well-being, conducting premature raids on the
    gangsters."
    (S. HTTP://Gloucester.Insomnia247.NL/ fuer Kontaktdaten!)
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Tue Nov 10 07:47:53 2026
    From Newsgroup: sci.electronics.design

    On 11/10/2026 7:01 AM, C||il|!n Niocl|is|!n Glost|-ir wrote:
    Johann 'Myrkraverk' Oskarsson <johann@Myrkraverk.invalid> skrev: |----------------------------------------------------------------------|
    |"[. . .] | |What is not amusing is that nobody thought to put computer security | |education in the elementary schools fifty years ago." | |----------------------------------------------------------------------|

    It is not plausible that a computer-security education in the
    elementary schools fifty years ago (in 1976) would be useful. Almost
    no child in 1976 used to have an Apple I, and an Apple I does not
    demand a password. Cf.
    HTTP://WWW.oldComputers.net

    Johann 'Myrkraverk' Oskarsson <johann@Myrkraverk.invalid> skrev: |----------------------------------------------------------------------|
    |"[. . .] |
    |[. . .] I don't know what goes on in the heads of peo- |
    |ple who decide the elementary curriculum. Usefulness in life is not | |their agenda. |
    | |
    |[. . .]" | |----------------------------------------------------------------------|

    Of course it is! Elementary schools teach useful mathematics and
    useful words.

    Johann 'Myrkraverk' Oskarsson <johann@Myrkraverk.invalid> skrev: |----------------------------------------------------------------------| |"Err, for an introduction to that, you could use the book /Undercover | |Cop/ by Mike Russel and Patrick W. Picciarelli. It's an autobiography|
    |of sorts, and therefore describes the real world, and not the fantasy | |world in movies. [. . .] |
    |[. . .] |
    | |
    |[. . .]" | |----------------------------------------------------------------------|

    Thanks for this suggestion. How did Mike Russell survive "being
    ambushed and shot in the back of the head"?

    Is the next quotation true? "his police handlers disregarded his
    wishes and his well-being, conducting premature raids on the
    gangsters."
    (S. HTTP://Gloucester.Insomnia247.NL/ fuer Kontaktdaten!)

    I don't recall the book exactly, but yes, that sounds true. Part of the
    theme of the book is the incompetence on the part of the police team --
    this includes Mike himself -- in planning and executing the undercover
    stint. But then, it's not like organized criminals are known to be
    really smart.

    As for how he survived, it's best you just read the book to find out.

    If I recall correctly, and then the TV shows correctly, this book fills
    in a lot of the gaps left implied in /The Sopranos/.
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Tue Nov 10 07:49:58 2026
    From Newsgroup: sci.electronics.design

    On 11/10/2026 6:36 AM, C||il|!n Niocl|is|!n Glost|-ir wrote:
    Books and phones can become stolen.
    (S. HTTP://Gloucester.Insomnia247.NL/ fuer Kontaktdaten!)

    Yes, with the latter being more commonly stolen than the former. What
    do you do if your 2FA application isn't duplicated on your backup phone?

    You /do/ have a backup phone, right?
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From john larkin@jl@glen--canyon.com to sci.electronics.design on Mon Nov 9 19:23:44 2026
    From Newsgroup: sci.electronics.design

    On Mon, 9 Nov 2026 23:31:54 +0800, Johann 'Myrkraverk' Oskarsson <johann@myrkraverk.invalid> wrote:

    On 10/31/2026 11:17 PM, Bill Sloman wrote:

    I don't make a habit of it, but there's no diplomatic way of getting the
    message across, when people are ignorant (which happens quite a lot), or
    stupid (which is rarer).

    Of course, you never see the same fault in yourself, you /fucking
    moron/!

    It's not anyone's fault that they are stupid. 50% of the population is
    below the median IQ.

    Be kind to the people who weren't lucky to be born beautiful geniuses
    like us.

    It is people's fault if they are nasty.



    John Larkin
    Highland Tech Glen Canyon Design Center
    Lunatic Fringe Electronics
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Tue Nov 10 16:10:31 2026
    From Newsgroup: sci.electronics.design

    On 10/11/2026 2:31 am, Johann 'Myrkraverk' Oskarsson wrote:
    On 10/31/2026 11:17 PM, Bill Sloman wrote:

    I don't make a habit of it, but there's no diplomatic way of getting
    the message across, when people are ignorant (which happens quite a
    lot), or stupid (which is rarer).

    Of course, you never see the same fault in yourself, you /fucking
    moron/!

    There are loads of areas where I'm totally ignorant. Real world feedback suggests that I'm not stupid, but in areas where you are ignorant you
    can end up looking pretty stupid to people who are less ignorant.

    Asserting that somebody is a /fucking moron/ is simple abuse, and that
    does seem to be the only kind that you can manage.
    --
    Bill Sloman, Sydney
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Tue Nov 10 16:18:38 2026
    From Newsgroup: sci.electronics.design

    On 10/11/2026 2:23 pm, john larkin wrote:
    On Mon, 9 Nov 2026 23:31:54 +0800, Johann 'Myrkraverk' Oskarsson <johann@myrkraverk.invalid> wrote:

    On 10/31/2026 11:17 PM, Bill Sloman wrote:

    I don't make a habit of it, but there's no diplomatic way of getting the >>> message across, when people are ignorant (which happens quite a lot), or >>> stupid (which is rarer).

    Of course, you never see the same fault in yourself, you /fucking
    moron/!

    It's not anyone's fault that they are stupid. 50% of the population is
    below the median IQ.

    I'm not. It's hard to get a Ph.D. if you can't score higher than 115 on
    a IQ test, and I didn't have any trouble doing that, though I did take
    a while to finish the project. The IQ test is a cheap, simple and none
    too reliable measure of intelligence, but stupid people do pay more
    attention to IQ scores than they should.

    Be kind to the people who weren't lucky to be born beautiful geniuses
    like us.

    John Larkin isn't a genius - neither is Donald Trump - but egomaniacs do
    like to make those kinds of claims

    It is people's fault if they are nasty.

    In some situations being nasty is the only way to get the message across.
    --
    Bill Sloman, Sydney
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Tue Nov 10 16:32:03 2026
    From Newsgroup: sci.electronics.design

    On 10/11/2026 2:40 am, Johann 'Myrkraverk' Oskarsson wrote:
    On 10/30/2026 9:36 PM, Martin Brown wrote:
    I was always struck by the typical senior managers password being so
    weak that it was hardly worth having. Some of them even had it written
    on a postit and stuck to their screen! You couldn't make it up!

    By observation -- and this is something Bill Sloman never understands,
    he doesn't observe anything -

    I wonder what observation Johann 'Myrkraverk' Oskarsson based this clain
    on? It looks more like mindless abuse.

    - the fault lies with nobody of sufficient
    social status being around to /correct/ the managers.

    Managers don't like being corrected, and they do tend to manage their environments so that it doesn't happen often, and never in public.
    > And since they are the ones hiring and firing the /security consultants/
    their own view is more important than whoever tries to teach them good password habits.

    Really, it's fine to write passwords down; just do it in a book you keep
    in your pocket, and don't forget it by the computer [terminal].

    And once you start writing the passwords down, you don't need to be
    afraid of the cryptographically secure random password generators.

    2FA helps if the password gets stolen somehow; it's usually not lifted
    from the book.-a It's still annoying.

    Two factor indentification for me is getting a text message on my mobile
    phone (which I do keep in my hip pocket). My bank has sent me a random
    number generator which puts up a new six digit random number every
    minute or so, and telling them what it is showing when they ask me seems
    to be a sufficient additional identifier for them.
    --
    Bill Sloman, Sydney

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Tue Nov 10 16:49:55 2026
    From Newsgroup: sci.electronics.design

    On 10/11/2026 2:12 am, john larkin wrote:
    On Mon, 9 Nov 2026 16:42:08 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 9/11/2026 6:27 am, Waldek Hebisch wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:
    On 7/11/2026 3:57 pm, john larkin wrote:
    On Sat, 7 Nov 2026 15:38:13 +1100, Bill Sloman <bill.sloman@ieee.org> >>>>> wrote:

    On 7/11/2026 5:54 am, john larkin wrote:
    On Fri, 6 Nov 2026 10:24:07 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and data. Signed drivers

    Not a Lisp fan then?

    People love to invent weird computer languages. There are over 3000 >>>>>>>>> and they keep coming.

    Real men code in binary using the front panel switches.

    I actually used a PDP-7 where you really did have to bootstrap it that >>>>>>>> way to read in a paper tape. It was a fun machine as it had a vector >>>>>>>> graphics display so was very good for asteroids, but cheated at chess. >>>>>>>
    I did a lot of stuff in PDP-11 and 68K assembly languages. With a few >>>>>>> macros, they are almost higher-level languages.

    All high level languages get compiled into assembly language before they >>>>>> can be run on a computer.

    Well, no.

    Counter-example?

    Being a compiler is tedious.

    Unlike some architectures that I might name.

    A language isn't an architecture. It has to be processed into something >>>>>> that can be run on hardware (which will have an architecture - it is >>>>>> constructed from components that have been connected by some kind of >>>>>> architect).

    If that post were compiled to assembly language, it would be a nop.

    As a compiler, you do seem to be sub-optimal, if not downright
    afunctional. Ignoring implications you don't like isn't an option for a >>>> compiler. There can be functions they can't parse - in your case don't >>>> want to parse - but that doesn't generate a nop.

    You clearly go into an area where you are not an expert.

    I'm not pretending to be an expert, just less of a clueless newbie in
    the area that John Larkin has shown himself to be.

    I've written several compilers and several RTOS's and a bunch of
    system simulators and hundreds of application programs, a good chunk
    of a million lines of code. From PDP8 to ARM and one CPU that I
    designed myself.

    And I'm sure that you are perfectly happy with all of them. Egomaniacs
    do go around describing themself as "stable geniuses".
    What's your definition of newbie?

    Somebody who doesn't know enough about what he his doing to be aware of
    the areas he doesn't know enough about. It's not so much about what you
    know, as what you know about the limits of your knowledge.
    --
    Bill Sloman, Sydney
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From liz@liz@poppyrecords.invalid.invalid (Liz Tuddenham) to sci.electronics.design on Tue Nov 10 10:12:57 2026
    From Newsgroup: sci.electronics.design

    Bill Sloman <bill.sloman@ieee.org> wrote:

    [...]
    in areas where you are ignorant you
    can end up looking pretty stupid to people who are less ignorant.

    Conversely, in areas where you are knowledgeable, you can look stupid to
    people who are ignorant.
    --
    ~ Liz Tuddenham ~
    (Remove the ".invalid"s and add ".co.uk" to reply)
    www.poppyrecords.co.uk
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Tue Nov 10 18:20:58 2026
    From Newsgroup: sci.electronics.design

    On 11/10/2026 1:49 PM, Bill Sloman wrote:
    On 10/11/2026 2:12 am, john larkin wrote:
    On Mon, 9 Nov 2026 16:42:08 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 9/11/2026 6:27 am, Waldek Hebisch wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:
    On 7/11/2026 3:57 pm, john larkin wrote:
    On Sat, 7 Nov 2026 15:38:13 +1100, Bill Sloman <bill.sloman@ieee.org> >>>>>> wrote:

    On 7/11/2026 5:54 am, john larkin wrote:
    On Fri, 6 Nov 2026 10:24:07 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and data. >>>>>>>>>>>> Signed drivers

    Not a Lisp fan then?

    People love to invent weird computer languages. There are over >>>>>>>>>> 3000
    and they keep coming.

    Real men code in binary using the front panel switches.

    I actually used a PDP-7 where you really did have to bootstrap >>>>>>>>> it that
    way to read in a paper tape. It was a fun machine as it had a >>>>>>>>> vector
    graphics display so was very good for asteroids, but cheated at >>>>>>>>> chess.

    I did a lot of stuff in PDP-11 and 68K assembly languages. With >>>>>>>> a few
    macros, they are almost higher-level languages.

    All high level languages get compiled into assembly language
    before they
    can be run on a computer.

    Well, no.

    Counter-example?

    Being a compiler is tedious.

    Unlike some architectures that I might name.

    A language isn't an architecture. It has to be processed into
    something
    that can be run on hardware (which will have an architecture - it is >>>>>>> constructed from components that have been connected by some kind of >>>>>>> architect).

    If that post were compiled to assembly language, it would be a nop. >>>>>
    As a compiler, you do seem to be sub-optimal, if not downright
    afunctional. Ignoring implications you don't like isn't an option
    for a
    compiler. There can be functions they can't parse - in your case don't >>>>> want to parse - but that doesn't generate a nop.

    You clearly go into an area where you are not an expert.

    I'm not pretending to be an expert, just less of a clueless newbie in
    the area that John Larkin has shown himself to be.

    I've written several compilers and several RTOS's and a bunch of
    system simulators and hundreds of application programs, a good chunk
    of a million lines of code. From PDP8 to ARM and one CPU that I
    designed myself.

    And I'm sure that you are perfectly happy with all of them. Egomaniacs
    do go around describing themself as "stable geniuses".
    What's your definition of newbie?

    Somebody who doesn't know enough about what he his doing to be aware of
    the areas he doesn't know enough about. It's not so much about what you know, as what you know about the limits of your knowledge.


    And then we have the military term "unknown unknown." For instance in business, you cannot prevent someone from inventing something else, that completely obsoletes your business.

    In /tech/, this is not an unknown unknown, because it's been done so
    many times in living memory. It can still happen in other fields, com-
    pletely blind-sighting the billionaires in charge of it.

    Some people refer to this as a /black swan/ event, which I believe comes
    from the field of economics.

    And, to combine this topic with another one, since we were talking about compilers, in the 90s, prior to the publication of GCC 2.95 [1], it
    would have been considered an /unknown unknown/ that compilers could be-
    come malicious to security sensitive programming. I bring you [2] as
    exhibit A, and give the relevant quote.

    The short and long is that, today, the only way to access a constant
    time implementation from Rust is to write said implementation in
    assembly.

    I'll add that this is about comparing security sensitive strings, us-
    ually of fixed length [3], and /not/ an O( 1 ) algorithm.

    You can read this poster's point about why LLVM is not a polite com-
    piler at [2].

    Now, if you want to argue this conclusion, you'll have to make an acc-
    ount on Stack Overflow, and comment there.

    In other words, the only way to write constant time cryptography, is ba- sically to do it in assembly, if any of GCC, LLVM, or MSVC are to com-
    pile it.

    In that sense, the /social contract/ between security sensitive program-
    ming, and compiler makers has been broken.

    And since we are in sci.electronics.design, we can also add that hard-
    ware manufacturers typically do not assert in their manuals that a spe-
    cific instruction is constant time, so now we get to wait and see if any
    of the x86, or ARM, or maybe even RISC-V manufacturers bring back the
    evil that was -- to my recollection -- the original 68000 shift instruc-
    tion which spent more clock time for larger shift values.

    And to finish this off with a warning, I have read the above for adults,
    so if you less-than-jr-high-school reading comprehension isn't up to the
    task of figuring out what I'm saying, that's squarely on you.



    Best wishes, and happy constant time algorithms!


    [1] Given Anton Ertl's research.

    [2] https://stackoverflow.com/questions/44691363/how-to-compare-strings-in-constant-time

    [3] Hashed passwords are always the same length, but if the length
    changes, comparing them in constant time doesn't work; this should be
    obvious.
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Tue Nov 10 18:22:05 2026
    From Newsgroup: sci.electronics.design

    On 11/10/2026 6:20 PM, Johann 'Myrkraverk' Oskarsson wrote:

    And to finish this off with a warning, I have read the above for adults,
    so if you less-than-jr-high-school reading comprehension isn't up to the
    task of figuring out what I'm saying, that's squarely on you.
    Sorry, *written; I send before proof reading!
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Tue Nov 10 18:31:57 2026
    From Newsgroup: sci.electronics.design

    On 11/10/2026 1:10 PM, Bill Sloman wrote:
    On 10/11/2026 2:31 am, Johann 'Myrkraverk' Oskarsson wrote:
    On 10/31/2026 11:17 PM, Bill Sloman wrote:

    I don't make a habit of it, but there's no diplomatic way of getting
    the message across, when people are ignorant (which happens quite a
    lot), or stupid (which is rarer).

    Of course, you never see the same fault in yourself, you /fucking
    moron/!

    There are loads of areas where I'm totally ignorant. Real world feedback suggests that I'm not stupid, but in areas where you are ignorant you
    can end up looking pretty stupid to people who are less ignorant.

    Asserting that somebody is a /fucking moron/ is simple abuse, and that
    does seem to be the only kind that you can manage.


    No, absolutely not. I'm capable of quite complex abuse, which you'd re-
    member if you bothered to remember anything.

    The point here, is that you deserve the abuse. Everyone here knows it,
    and now you do too. You seem to think that throwing abusing in some-
    one's face doesn't get you abuse thrown in yours. Now I'm here to edu-
    cate you on the error of your ways. The education you should have gott-
    en in kindergarten but didn't. Maybe you didn't get to kindergarten.

    You see, you seem to be completely ignorant of how the real world works.
    So now you get to /feel/ the consequences of your own actions, of origi-
    nally throwing abuse in my direction. You missed. And now you get more
    back.


    Happy receiving abuse, you fucking moron!
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Tue Nov 10 18:38:47 2026
    From Newsgroup: sci.electronics.design

    On 11/10/2026 6:12 PM, Liz Tuddenham wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:

    [...]
    in areas where you are ignorant you
    can end up looking pretty stupid to people who are less ignorant.

    Conversely, in areas where you are knowledgeable, you can look stupid to people who are ignorant.


    And this happens all the time. It's kind of better to be a /midwit/
    than a genius [1]. I've seen that a lot, in online /special interest/
    groups, and be subject to such abuse myself. The problem when people
    get abusive, and say someone is /incompetent/, is that they are gener-
    ally wrong. First, the moron can learn, and become greater that the
    person throwing the abuse. This is an effect of time. Second, the ab-
    user rarely follows up with /why/ said abuse was thrown, so the receiver
    is left to ponder -- is there something I don't know, or are you just a
    moron?

    Now, people been around the block a few times know that today's ignora-
    mus can be a /jedi master/ in a few months, or years, and doesn't have
    to tell the rest of the group about their knowledge accumulation. So
    they always end up as /just another moron/ once the person knows better.

    This is called /lack of social skills/, and has real consequences.
    Something some members of sci.electronics.design have yet to learn.


    Best wishes, and happy jedi mastery!


    [1] If we use that term, the meme with the IQ curve I'm referring to has
    a /moron/ on the left, and a /jedi master/ on the right. The terms do
    not matter so much as long as we're being understood.
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Tue Nov 10 18:53:39 2026
    From Newsgroup: sci.electronics.design

    On 11/10/2026 2:23 AM, Don Y wrote:
    On 11/9/2026 8:28 AM, Johann 'Myrkraverk' Oskarsson wrote:
    On 10/29/2026 8:59 PM, Don Y wrote:
    Management of these processes is the cause of most bad designs.
    The folks doing so are usually pretty ignorant of the efforts they
    are charged with managing!

    Please excuse me, I probably showed this example earlier this year in
    sci.electronics.design.-a A good example is

    -a-a https://neilmadden.blog/2022/04/19/psychic-signatures-in-java/

    where the /management/ was too lazy to hire a cryptographer for the pro-
    ject.-a It's not like you can accuse Oracle of not affording one.

    This (and the opposite:-a hiring someone and then ignoring their "expert recommendations") happens surprisingly often.

    The people making the decisions aren't aware of how much they are unaware. Dunning-Kruger.-a Particularly dangerous when The Ignorant are in charge.

    Nowhere is this more evident than in software as ANYONE can write code
    and, true to D-K, they all *think* they know far more than they actually do.-a "I've written..." yeah, but were any of those really more than trivial exercises?-a I.e., if it was the sort of problem we'd get in a homework assignment, you know it has a complexity on the order of 8 hours or less
    (cuz the homework budget for a course was 8 hours)

    Another favorite is "real-time" -- particularly "hard" vs. "soft".
    When most folks think it means "real fast" (it doesn't).-a Or, that
    "hard" is "more difficult" than "soft" (the opposite is true -- in
    spades -- which is why so many newbies resort to HRT solutions
    to problems... they can't think beyond FORCING a hard requirement
    when none is required... even in nuclear defenses!)

    But, so much of this is ingrained that challenging it threatens
    most folks' notion of their own competence so they ignore learning
    about those things where their ignorance is so profound.

    [You have to learn to work AROUND such people lest they compromise
    any good work]

    Thank you. When it comes to highly paid managers, or CEOs for that
    matter, I have this theory that /whoever is paid more is right/. Which
    while watching financial crisis reenactment movies [1] seems to be true.

    Of course I'm not completely sure of this, and there are always excep-
    tions, so we can consider it a /working theory/, and add more observa-
    tions of managers to the mix to see if it comes out statistically true
    or false.

    On the prior topic, I also lack experience in hard real time design, and
    don't even have a lot to say about the soft part of it. It's something
    I'll keep in mind whenever I find myself facing such projects.


    Best wishes, and happy real time projects!

    [1] They are never absolutely true to life, so need to be taken with a
    pillar of salt, or maybe several.
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Tue Nov 10 19:02:16 2026
    From Newsgroup: sci.electronics.design

    On 11/10/2026 5:14 AM, Waldek Hebisch wrote:
    Actually, at '-O2' version of 'gcc' that I use replaces explicit
    division by 0 by the 'ud2' instruction.

    Shouldn't that be considered /erroneous compilation/? I mean, you can
    write a signal handler that catches all signals, and isn't /division by
    zero/ one of the Posix signals? Is the compiler prescient and makes
    sure this only happens if the binary is ever run when the user isn't ex- plictly expecting a division by zero -- rather than /invalid opcode/?

    Does the question bother you? Am I too high, or too low on the /moron/
    vs. /jedi master/ axis to be understood? Should we put all the GCC devs
    in therapy for believing they are prescient?
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Tue Nov 10 23:41:49 2026
    From Newsgroup: sci.electronics.design

    On 10/11/2026 9:20 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 1:49 PM, Bill Sloman wrote:
    On 10/11/2026 2:12 am, john larkin wrote:
    On Mon, 9 Nov 2026 16:42:08 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 9/11/2026 6:27 am, Waldek Hebisch wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:
    On 7/11/2026 3:57 pm, john larkin wrote:
    On Sat, 7 Nov 2026 15:38:13 +1100, Bill Sloman
    <bill.sloman@ieee.org>
    wrote:

    On 7/11/2026 5:54 am, john larkin wrote:
    On Fri, 6 Nov 2026 10:24:07 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and data. >>>>>>>>>>>>> Signed drivers

    Not a Lisp fan then?

    People love to invent weird computer languages. There are >>>>>>>>>>> over 3000
    and they keep coming.

    Real men code in binary using the front panel switches.

    I actually used a PDP-7 where you really did have to bootstrap >>>>>>>>>> it that
    way to read in a paper tape. It was a fun machine as it had a >>>>>>>>>> vector
    graphics display so was very good for asteroids, but cheated >>>>>>>>>> at chess.

    I did a lot of stuff in PDP-11 and 68K assembly languages. With >>>>>>>>> a few
    macros, they are almost higher-level languages.

    All high level languages get compiled into assembly language
    before they
    can be run on a computer.

    Well, no.

    Counter-example?

    Being a compiler is tedious.

    Unlike some architectures that I might name.

    A language isn't an architecture. It has to be processed into >>>>>>>> something
    that can be run on hardware (which will have an architecture - >>>>>>>> it is
    constructed from components that have been connected by some
    kind of
    architect).

    If that post were compiled to assembly language, it would be a nop. >>>>>>
    As a compiler, you do seem to be sub-optimal, if not downright
    afunctional. Ignoring implications you don't like isn't an option >>>>>> for a
    compiler. There can be functions they can't parse - in your case
    don't
    want to parse - but that doesn't generate a nop.

    You clearly go into an area where you are not an expert.

    I'm not pretending to be an expert, just less of a clueless newbie in
    the area that John Larkin has shown himself to be.

    I've written several compilers and several RTOS's and a bunch of
    system simulators and hundreds of application programs, a good chunk
    of a million lines of code. From PDP8 to ARM and one CPU that I
    designed myself.

    And I'm sure that you are perfectly happy with all of them. Egomaniacs
    do go around describing themself as "stable geniuses".
    What's your definition of newbie?

    Somebody who doesn't know enough about what he his doing to be aware
    of the areas he doesn't know enough about. It's not so much about what
    you know, as what you know about the limits of your knowledge.

    <snipped pretentious twaddle>
    --
    Bill Sloman, Sydney

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Tue Nov 10 23:46:51 2026
    From Newsgroup: sci.electronics.design

    On 10/11/2026 9:22 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 6:20 PM, Johann 'Myrkraverk' Oskarsson wrote:

    And to finish this off with a warning, I have read the above for adults,
    so if you less-than-jr-high-school reading comprehension isn't up to the
    task of figuring out what I'm saying, that's squarely on you.
    Sorry, *written; I send before proof reading!

    "Sent" before proof reading. You imagine that you wrote your pretentious twaddle for adults, not realising that the "adults" have noticed that
    you don't have anything useful to say, and have mostly got around to
    plonking you.
    --
    Bill Sloman, Sydney
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Wed Nov 11 00:08:55 2026
    From Newsgroup: sci.electronics.design

    On 10/11/2026 9:38 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 6:12 PM, Liz Tuddenham wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:

    [1] If we use that term, the meme with the IQ curve I'm referring to has
    a /moron/ on the left, and a /jedi master/ on the right.-a The terms do
    not matter so much as long as we're being understood.

    What's obvious that you understand IQ curves to be scoring a
    single-valued parameter.

    It's well known that there are lots of components to the faculty that is usually described as intelligence. Quite a lot of work was done early on
    om multifactor intelligence tests, but it didn't go anywhere useful.

    The people doing the work didn't have much of a clue what the different factors might be, and didn't find any practical way of teasing out what
    they might be. When people in recent years started using magnetic
    resonance images of the brain to work out what was going on in the
    brain, where, and in what sequence, they didn't do any better.

    It's complicated and we need tools with a lot better discrimination than
    we have at the moment.

    If you think that you can divide the world into a linear array of people running from moron to the left to jedi master on the right, you've put yourself into the moron category.

    My late wife was a psycholinguist who spoke four languages fluently and
    was a foreign member of the US Academy of Science and a Fellow of the
    Royal Society of London, but she didn't know much about electronics. At
    least she had enough sense not to post here.
    --
    Bill Sloman, Sydney
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Tue Nov 10 21:30:32 2026
    From Newsgroup: sci.electronics.design

    On 11/10/2026 8:46 PM, Bill Sloman wrote:
    On 10/11/2026 9:22 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 6:20 PM, Johann 'Myrkraverk' Oskarsson wrote:

    And to finish this off with a warning, I have read the above for adults, >>> so if you less-than-jr-high-school reading comprehension isn't up to the >>> task of figuring out what I'm saying, that's squarely on you.
    Sorry, *written; I send before proof reading!

    "Sent" before proof reading. You imagine that you wrote your pretentious twaddle for adults, not realising that the "adults" have noticed that
    you don't have anything useful to say, and have mostly got around to plonking you.


    Speaking for yourself, are you? Because here in sci.electronics.design,
    you're the only one plonking. And you're plonking everyone else.

    So you're the bully who has never faced the consequences of his own ac-
    tions.

    And I'm here now, and I'll take none of it silently. You are a fucking
    bully, and you don't know it.

    Now you do. And everyone else here in sci.electronics.design knew it
    too but was too polite to tell you -- maybe. I didn't read the 30 year backlog.


    Best wishes, and keep being miserable for the next fifty years -- at
    least!
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Wed Nov 11 00:31:34 2026
    From Newsgroup: sci.electronics.design

    On 10/11/2026 9:31 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 1:10 PM, Bill Sloman wrote:
    On 10/11/2026 2:31 am, Johann 'Myrkraverk' Oskarsson wrote:
    On 10/31/2026 11:17 PM, Bill Sloman wrote:

    I don't make a habit of it, but there's no diplomatic way of getting
    the message across, when people are ignorant (which happens quite a
    lot), or stupid (which is rarer).

    Of course, you never see the same fault in yourself, you /fucking
    moron/!

    There are loads of areas where I'm totally ignorant. Real world
    feedback suggests that I'm not stupid, but in areas where you are
    ignorant you can end up looking pretty stupid to people who are less
    ignorant.

    Asserting that somebody is a /fucking moron/ is simple abuse, and that
    does seem to be the only kind that you can manage.


    No, absolutely not.-a I'm capable of quite complex abuse, which you'd re- member if you bothered to remember anything.

    Really?

    <snip>

    Happy receiving abuse, you fucking moron!

    I'd prefer that you didn't waste bandwidth posting pointless drivel, but
    I really couldn't care less.
    --
    Bill Sloman, Sydney

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Tue Nov 10 21:40:37 2026
    From Newsgroup: sci.electronics.design

    On 11/10/2026 9:08 PM, Bill Sloman wrote:
    On 10/11/2026 9:38 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 6:12 PM, Liz Tuddenham wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:

    [1] If we use that term, the meme with the IQ curve I'm referring to has
    a /moron/ on the left, and a /jedi master/ on the right.-a The terms do
    not matter so much as long as we're being understood.

    What's obvious that you understand IQ curves to be scoring a single-
    valued parameter.

    It's well known that there are lots of components to the faculty that is usually described as intelligence. Quite a lot of work was done early on
    om multifactor intelligence tests, but it didn't go anywhere useful.

    The people doing the work didn't have much of a clue what the different factors might be, and didn't find any practical way of teasing out what
    they might be. When people in recent years started using magnetic
    resonance images of the brain to work out what was going on in the
    brain, where, and in what sequence, they didn't do any better.

    It's complicated and we need tools with a lot better discrimination than
    we have at the moment.

    If you think that you can divide the world into a linear array of people running from moron to the left to jedi master on the right, you've put yourself into the moron category.

    My late wife was a psycholinguist who spoke four languages fluently and
    was a foreign member of the US Academy of Science and a Fellow of the
    Royal Society of London, but she didn't know much about electronics. At least she had enough sense not to post here.


    Well, I never met your late wife, but I can tell you that I don't be-
    lieve anything about the current "understanding" of IQ. It's only use-
    ful in one context, and that's measuring learning disability. Anything
    else is just handwavium and useless twiddle twaddle among the academi-
    cians.

    This belief is based on actual /real life/ experience, something you
    have never experienced yourself, demonstrating time and again that you
    don't know how life works.

    And, no; I'm not going to bother explaining anything to you, because
    this is outside the scope of sci.electronics.design, and the rest of
    the people aren't asking. /You/ personally are not worth explaining
    anything to. You know that, don't you?


    Best wishes, and be even more miserable than usual!
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Wed Nov 11 00:41:11 2026
    From Newsgroup: sci.electronics.design

    On 10/11/2026 9:53 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 2:23 AM, Don Y wrote:
    On 11/9/2026 8:28 AM, Johann 'Myrkraverk' Oskarsson wrote:
    On 10/29/2026 8:59 PM, Don Y wrote:
    Management of these processes is the cause of most bad designs.
    The folks doing so are usually pretty ignorant of the efforts they
    are charged with managing!

    Please excuse me, I probably showed this example earlier this year in
    sci.electronics.design.-a A good example is

    -a-a https://neilmadden.blog/2022/04/19/psychic-signatures-in-java/

    where the /management/ was too lazy to hire a cryptographer for the pro- >>> ject.-a It's not like you can accuse Oracle of not affording one.

    This (and the opposite:-a hiring someone and then ignoring their "expert
    recommendations") happens surprisingly often.

    The people making the decisions aren't aware of how much they are
    unaware.
    Dunning-Kruger.-a Particularly dangerous when The Ignorant are in charge.

    Nowhere is this more evident than in software as ANYONE can write code
    and, true to D-K, they all *think* they know far more than they actually
    do.-a "I've written..." yeah, but were any of those really more than
    trivial
    exercises?-a I.e., if it was the sort of problem we'd get in a homework
    assignment, you know it has a complexity on the order of 8 hours or less
    (cuz the homework budget for a course was 8 hours)

    Another favorite is "real-time" -- particularly "hard" vs. "soft".
    When most folks think it means "real fast" (it doesn't).-a Or, that
    "hard" is "more difficult" than "soft" (the opposite is true -- in
    spades -- which is why so many newbies resort to HRT solutions
    to problems... they can't think beyond FORCING a hard requirement
    when none is required... even in nuclear defenses!)

    But, so much of this is ingrained that challenging it threatens
    most folks' notion of their own competence so they ignore learning
    about those things where their ignorance is so profound.

    [You have to learn to work AROUND such people lest they compromise
    any good work]

    Thank you.-a When it comes to highly paid managers, or CEOs for that
    matter, I have this theory that /whoever is paid more is right/.-a Which while watching financial crisis reenactment movies [1] seems to be true.

    This would mean that American managers, who are paid huge multiples of
    what the median employee in their firm gets, are right much more often
    than Japaneses managers who aren't as well paid.

    Of course I'm not completely sure of this, and there are always excep-
    tions, so we can consider it a /working theory/, and add more observa-
    tions of managers to the mix to see if it comes out statistically true
    or false.

    You don't seem to have made even the most basic observation required to
    test the hypothesis.

    On the prior topic, I also lack experience in hard real time design, and don't even have a lot to say about the soft part of it.-a It's something
    I'll keep in mind whenever I find myself facing such projects.

    Any manager worth their salt wouldn't put you on such a project if they
    had any choice.
    --
    Bill Sloman, Sydney

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From john larkin@jl@glen--canyon.com to sci.electronics.design on Tue Nov 10 07:37:47 2026
    From Newsgroup: sci.electronics.design

    On Tue, 10 Nov 2026 16:18:38 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 10/11/2026 2:23 pm, john larkin wrote:
    On Mon, 9 Nov 2026 23:31:54 +0800, Johann 'Myrkraverk' Oskarsson
    <johann@myrkraverk.invalid> wrote:

    On 10/31/2026 11:17 PM, Bill Sloman wrote:

    I don't make a habit of it, but there's no diplomatic way of getting the >>>> message across, when people are ignorant (which happens quite a lot), or >>>> stupid (which is rarer).

    Of course, you never see the same fault in yourself, you /fucking
    moron/!

    It's not anyone's fault that they are stupid. 50% of the population is
    below the median IQ.

    I'm not. It's hard to get a Ph.D. if you can't score higher than 115 on
    a IQ test, and I didn't have any trouble doing that, though I did take
    a while to finish the project. The IQ test is a cheap, simple and none
    too reliable measure of intelligence, but stupid people do pay more >attention to IQ scores than they should.

    Be kind to the people who weren't lucky to be born beautiful geniuses
    like us.

    John Larkin isn't a genius - neither is Donald Trump - but egomaniacs do >like to make those kinds of claims

    It is people's fault if they are nasty.

    In some situations being nasty is the only way to get the message across.

    It's the stupid, nasty, lazy, least effective way.

    Especially when you're not right.

    I think that's recursive.


    John Larkin
    Highland Tech Glen Canyon Design Center
    Lunatic Fringe Electronics
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From john larkin@jl@glen--canyon.com to sci.electronics.design on Tue Nov 10 07:42:13 2026
    From Newsgroup: sci.electronics.design

    On Tue, 10 Nov 2026 18:38:47 +0800, Johann 'Myrkraverk' Oskarsson <johann@myrkraverk.invalid> wrote:

    On 11/10/2026 6:12 PM, Liz Tuddenham wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:

    [...]
    in areas where you are ignorant you
    can end up looking pretty stupid to people who are less ignorant.

    Conversely, in areas where you are knowledgeable, you can look stupid to
    people who are ignorant.


    And this happens all the time. It's kind of better to be a /midwit/
    than a genius [1]. I've seen that a lot, in online /special interest/ >groups, and be subject to such abuse myself. The problem when people
    get abusive, and say someone is /incompetent/, is that they are gener-
    ally wrong. First, the moron can learn, and become greater that the
    person throwing the abuse. This is an effect of time. Second, the ab-
    user rarely follows up with /why/ said abuse was thrown, so the receiver
    is left to ponder -- is there something I don't know, or are you just a >moron?

    Now, people been around the block a few times know that today's ignora-
    mus can be a /jedi master/ in a few months, or years, and doesn't have
    to tell the rest of the group about their knowledge accumulation. So
    they always end up as /just another moron/ once the person knows better.

    This is called /lack of social skills/, and has real consequences.
    Something some members of sci.electronics.design have yet to learn.


    Best wishes, and happy jedi mastery!


    [1] If we use that term, the meme with the IQ curve I'm referring to has
    a /moron/ on the left, and a /jedi master/ on the right. The terms do
    not matter so much as long as we're being understood.

    Some people have extreme talents. It might be math, music, tennis,
    whatever. Some part of their brain is six-sigma good, often at the
    expense of other parts.

    The specialized-genius bit can also make them obnoxious.


    John Larkin
    Highland Tech Glen Canyon Design Center
    Lunatic Fringe Electronics
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From john larkin@jl@glen--canyon.com to sci.electronics.design on Tue Nov 10 07:53:49 2026
    From Newsgroup: sci.electronics.design

    On Tue, 10 Nov 2026 16:49:55 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 10/11/2026 2:12 am, john larkin wrote:
    On Mon, 9 Nov 2026 16:42:08 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 9/11/2026 6:27 am, Waldek Hebisch wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:
    On 7/11/2026 3:57 pm, john larkin wrote:
    On Sat, 7 Nov 2026 15:38:13 +1100, Bill Sloman <bill.sloman@ieee.org> >>>>>> wrote:

    On 7/11/2026 5:54 am, john larkin wrote:
    On Fri, 6 Nov 2026 10:24:07 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and data. Signed drivers

    Not a Lisp fan then?

    People love to invent weird computer languages. There are over 3000 >>>>>>>>>> and they keep coming.

    Real men code in binary using the front panel switches.

    I actually used a PDP-7 where you really did have to bootstrap it that
    way to read in a paper tape. It was a fun machine as it had a vector >>>>>>>>> graphics display so was very good for asteroids, but cheated at chess.

    I did a lot of stuff in PDP-11 and 68K assembly languages. With a few >>>>>>>> macros, they are almost higher-level languages.

    All high level languages get compiled into assembly language before they
    can be run on a computer.

    Well, no.

    Counter-example?

    Being a compiler is tedious.

    Unlike some architectures that I might name.

    A language isn't an architecture. It has to be processed into something >>>>>>> that can be run on hardware (which will have an architecture - it is >>>>>>> constructed from components that have been connected by some kind of >>>>>>> architect).

    If that post were compiled to assembly language, it would be a nop. >>>>>
    As a compiler, you do seem to be sub-optimal, if not downright
    afunctional. Ignoring implications you don't like isn't an option for a >>>>> compiler. There can be functions they can't parse - in your case don't >>>>> want to parse - but that doesn't generate a nop.

    You clearly go into an area where you are not an expert.

    I'm not pretending to be an expert, just less of a clueless newbie in
    the area that John Larkin has shown himself to be.

    I've written several compilers and several RTOS's and a bunch of
    system simulators and hundreds of application programs, a good chunk
    of a million lines of code. From PDP8 to ARM and one CPU that I
    designed myself.

    And I'm sure that you are perfectly happy with all of them. Egomaniacs
    do go around describing themself as "stable geniuses".

    I wouldn't do that!

    What's your definition of newbie?

    Somebody who doesn't know enough about what he his doing to be aware of
    the areas he doesn't know enough about. It's not so much about what you >know, as what you know about the limits of your knowledge.

    Electronic design talent is reality-checked by what works.

    This worked first build. One of my guys wants to change one resistor,
    so I guess I'll let him.

    https://www.dropbox.com/scl/fi/69wwpsoc0o47cidxwust7/B360_20260813_110116.jpg?rlkey=y0sxkzswmgj3wa9lvf9ihbdio&raw=1

    Actually, we may change the big yellow polyfuse too. Manufacturing
    wants to kink the leads for solderability, but that lifts it up and
    then it hits the top cover and then we'd have to bend it over and that
    would look goofy.

    Life sucks.


    John Larkin
    Highland Tech Glen Canyon Design Center
    Lunatic Fringe Electronics
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From john larkin@jl@glen--canyon.com to sci.electronics.design on Tue Nov 10 07:59:49 2026
    From Newsgroup: sci.electronics.design

    On Tue, 10 Nov 2026 18:20:58 +0800, Johann 'Myrkraverk' Oskarsson <johann@myrkraverk.invalid> wrote:

    On 11/10/2026 1:49 PM, Bill Sloman wrote:
    On 10/11/2026 2:12 am, john larkin wrote:
    On Mon, 9 Nov 2026 16:42:08 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 9/11/2026 6:27 am, Waldek Hebisch wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:
    On 7/11/2026 3:57 pm, john larkin wrote:
    On Sat, 7 Nov 2026 15:38:13 +1100, Bill Sloman <bill.sloman@ieee.org> >>>>>>> wrote:

    On 7/11/2026 5:54 am, john larkin wrote:
    On Fri, 6 Nov 2026 10:24:07 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and data. >>>>>>>>>>>>> Signed drivers

    Not a Lisp fan then?

    People love to invent weird computer languages. There are over >>>>>>>>>>> 3000
    and they keep coming.

    Real men code in binary using the front panel switches.

    I actually used a PDP-7 where you really did have to bootstrap >>>>>>>>>> it that
    way to read in a paper tape. It was a fun machine as it had a >>>>>>>>>> vector
    graphics display so was very good for asteroids, but cheated at >>>>>>>>>> chess.

    I did a lot of stuff in PDP-11 and 68K assembly languages. With >>>>>>>>> a few
    macros, they are almost higher-level languages.

    All high level languages get compiled into assembly language
    before they
    can be run on a computer.

    Well, no.

    Counter-example?

    Being a compiler is tedious.

    Unlike some architectures that I might name.

    A language isn't an architecture. It has to be processed into >>>>>>>> something
    that can be run on hardware (which will have an architecture - it is >>>>>>>> constructed from components that have been connected by some kind of >>>>>>>> architect).

    If that post were compiled to assembly language, it would be a nop. >>>>>>
    As a compiler, you do seem to be sub-optimal, if not downright
    afunctional. Ignoring implications you don't like isn't an option >>>>>> for a
    compiler. There can be functions they can't parse - in your case don't >>>>>> want to parse - but that doesn't generate a nop.

    You clearly go into an area where you are not an expert.

    I'm not pretending to be an expert, just less of a clueless newbie in
    the area that John Larkin has shown himself to be.

    I've written several compilers and several RTOS's and a bunch of
    system simulators and hundreds of application programs, a good chunk
    of a million lines of code. From PDP8 to ARM and one CPU that I
    designed myself.

    And I'm sure that you are perfectly happy with all of them. Egomaniacs
    do go around describing themself as "stable geniuses".
    What's your definition of newbie?

    Somebody who doesn't know enough about what he his doing to be aware of
    the areas he doesn't know enough about. It's not so much about what you
    know, as what you know about the limits of your knowledge.


    And then we have the military term "unknown unknown." For instance in >business, you cannot prevent someone from inventing something else, that >completely obsoletes your business.

    In /tech/, this is not an unknown unknown, because it's been done so
    many times in living memory. It can still happen in other fields, com- >pletely blind-sighting the billionaires in charge of it.

    Some people refer to this as a /black swan/ event, which I believe comes
    from the field of economics.

    And, to combine this topic with another one, since we were talking about >compilers, in the 90s, prior to the publication of GCC 2.95 [1], it
    would have been considered an /unknown unknown/ that compilers could be-
    come malicious to security sensitive programming. I bring you [2] as
    exhibit A, and give the relevant quote.

    The short and long is that, today, the only way to access a constant
    time implementation from Rust is to write said implementation in
    assembly.

    I'll add that this is about comparing security sensitive strings, us-
    ually of fixed length [3], and /not/ an O( 1 ) algorithm.

    You can read this poster's point about why LLVM is not a polite com-
    piler at [2].

    Now, if you want to argue this conclusion, you'll have to make an acc-
    ount on Stack Overflow, and comment there.

    In other words, the only way to write constant time cryptography, is ba- >sically to do it in assembly, if any of GCC, LLVM, or MSVC are to com-
    pile it.

    In that sense, the /social contract/ between security sensitive program- >ming, and compiler makers has been broken.

    And since we are in sci.electronics.design, we can also add that hard-
    ware manufacturers typically do not assert in their manuals that a spe-
    cific instruction is constant time, so now we get to wait and see if any
    of the x86, or ARM, or maybe even RISC-V manufacturers bring back the
    evil that was -- to my recollection -- the original 68000 shift instruc-
    tion which spent more clock time for larger shift values.

    And to finish this off with a warning, I have read the above for adults,
    so if you less-than-jr-high-school reading comprehension isn't up to the
    task of figuring out what I'm saying, that's squarely on you.



    Best wishes, and happy constant time algorithms!


    [1] Given Anton Ertl's research.

    [2] >https://stackoverflow.com/questions/44691363/how-to-compare-strings-in-constant-time

    [3] Hashed passwords are always the same length, but if the length
    changes, comparing them in constant time doesn't work; this should be >obvious.

    I'm always impressed by how few programmers have even a rough idea of
    how much time it takes to execute their code. Guessing off by 10:1 is
    common.

    I get a scope and measure it.

    We have a 75-cent dual-core ARM making a 100 MHz square wave, lines of
    c setting a port bit up and down.


    John Larkin
    Highland Tech Glen Canyon Design Center
    Lunatic Fringe Electronics
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From =?UTF-8?B?Q8OzaWzDrW4=?= =?UTF-8?B?IE5pb2Nsw6Fzw61u?= =?UTF-8?B?IEdsb3N0w6lpcg==?=@thanks-to@Taf.com to sci.electronics.design on Tue Nov 10 16:16:23 2026
    From Newsgroup: sci.electronics.design

    Johann 'Myrkraverk' Oskarsson <johann@Myrkraverk.invalid> skrev: |-----------------------------------------------------------------------|
    |"[. . .] Part of the | |theme of the book is the incompetence on the part of the police team --|
    |this includes Mike himself -- in planning and executing the undercover | |stint. [. . .] |
    |[. . .]" | |-----------------------------------------------------------------------|

    Hej!

    Do you know if there are any similar books about Scandinavian or
    Icelandic incompetent police forces or other incompetent authorities
    (e.g. courts)? (I do not need such books to be in English - I can read
    Swedish etc.) I asked a friend in Sweden is "Skiftet" on Netflix a
    true story. He did not know. (It is not a true story, but it is
    supposed to be realistic.) Thanks.

    |-----------------------------------------------------------------------|
    |"As for how he survived, it's best you just read the book to find out. |
    | |
    |[. . .]" | |-----------------------------------------------------------------------|

    The blurb is not convincing that a person could really survive being
    shot in his head. (Trump does survive, because an ear got shot.)
    (S. HTTP://Gloucester.Insomnia247.NL/ fuer Kontaktdaten!)
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Wed Nov 11 02:21:09 2026
    From Newsgroup: sci.electronics.design

    On 11/10/2026 11:59 PM, john larkin wrote:
    On Tue, 10 Nov 2026 18:20:58 +0800, Johann 'Myrkraverk' Oskarsson <johann@myrkraverk.invalid> wrote:

    On 11/10/2026 1:49 PM, Bill Sloman wrote:
    On 10/11/2026 2:12 am, john larkin wrote:
    On Mon, 9 Nov 2026 16:42:08 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 9/11/2026 6:27 am, Waldek Hebisch wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:
    On 7/11/2026 3:57 pm, john larkin wrote:
    On Sat, 7 Nov 2026 15:38:13 +1100, Bill Sloman <bill.sloman@ieee.org> >>>>>>>> wrote:

    On 7/11/2026 5:54 am, john larkin wrote:
    On Fri, 6 Nov 2026 10:24:07 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and data. >>>>>>>>>>>>>> Signed drivers

    Not a Lisp fan then?

    People love to invent weird computer languages. There are over >>>>>>>>>>>> 3000
    and they keep coming.

    Real men code in binary using the front panel switches.

    I actually used a PDP-7 where you really did have to bootstrap >>>>>>>>>>> it that
    way to read in a paper tape. It was a fun machine as it had a >>>>>>>>>>> vector
    graphics display so was very good for asteroids, but cheated at >>>>>>>>>>> chess.

    I did a lot of stuff in PDP-11 and 68K assembly languages. With >>>>>>>>>> a few
    macros, they are almost higher-level languages.

    All high level languages get compiled into assembly language >>>>>>>>> before they
    can be run on a computer.

    Well, no.

    Counter-example?

    Being a compiler is tedious.

    Unlike some architectures that I might name.

    A language isn't an architecture. It has to be processed into >>>>>>>>> something
    that can be run on hardware (which will have an architecture - it is >>>>>>>>> constructed from components that have been connected by some kind of >>>>>>>>> architect).

    If that post were compiled to assembly language, it would be a nop. >>>>>>>
    As a compiler, you do seem to be sub-optimal, if not downright
    afunctional. Ignoring implications you don't like isn't an option >>>>>>> for a
    compiler. There can be functions they can't parse - in your case don't >>>>>>> want to parse - but that doesn't generate a nop.

    You clearly go into an area where you are not an expert.

    I'm not pretending to be an expert, just less of a clueless newbie in >>>>> the area that John Larkin has shown himself to be.

    I've written several compilers and several RTOS's and a bunch of
    system simulators and hundreds of application programs, a good chunk
    of a million lines of code. From PDP8 to ARM and one CPU that I
    designed myself.

    And I'm sure that you are perfectly happy with all of them. Egomaniacs
    do go around describing themself as "stable geniuses".
    What's your definition of newbie?

    Somebody who doesn't know enough about what he his doing to be aware of
    the areas he doesn't know enough about. It's not so much about what you
    know, as what you know about the limits of your knowledge.


    And then we have the military term "unknown unknown." For instance in
    business, you cannot prevent someone from inventing something else, that
    completely obsoletes your business.

    In /tech/, this is not an unknown unknown, because it's been done so
    many times in living memory. It can still happen in other fields, com-
    pletely blind-sighting the billionaires in charge of it.

    Some people refer to this as a /black swan/ event, which I believe comes >>from the field of economics.

    And, to combine this topic with another one, since we were talking about
    compilers, in the 90s, prior to the publication of GCC 2.95 [1], it
    would have been considered an /unknown unknown/ that compilers could be-
    come malicious to security sensitive programming. I bring you [2] as
    exhibit A, and give the relevant quote.

    The short and long is that, today, the only way to access a constant
    time implementation from Rust is to write said implementation in
    assembly.

    I'll add that this is about comparing security sensitive strings, us-
    ually of fixed length [3], and /not/ an O( 1 ) algorithm.

    You can read this poster's point about why LLVM is not a polite com-
    piler at [2].

    Now, if you want to argue this conclusion, you'll have to make an acc-
    ount on Stack Overflow, and comment there.

    In other words, the only way to write constant time cryptography, is ba-
    sically to do it in assembly, if any of GCC, LLVM, or MSVC are to com-
    pile it.

    In that sense, the /social contract/ between security sensitive program-
    ming, and compiler makers has been broken.

    And since we are in sci.electronics.design, we can also add that hard-
    ware manufacturers typically do not assert in their manuals that a spe-
    cific instruction is constant time, so now we get to wait and see if any
    of the x86, or ARM, or maybe even RISC-V manufacturers bring back the
    evil that was -- to my recollection -- the original 68000 shift instruc-
    tion which spent more clock time for larger shift values.

    And to finish this off with a warning, I have read the above for adults,
    so if you less-than-jr-high-school reading comprehension isn't up to the
    task of figuring out what I'm saying, that's squarely on you.



    Best wishes, and happy constant time algorithms!


    [1] Given Anton Ertl's research.

    [2]
    https://stackoverflow.com/questions/44691363/how-to-compare-strings-in-constant-time

    [3] Hashed passwords are always the same length, but if the length
    changes, comparing them in constant time doesn't work; this should be
    obvious.

    I'm always impressed by how few programmers have even a rough idea of
    how much time it takes to execute their code. Guessing off by 10:1 is
    common.

    I get a scope and measure it.

    We have a 75-cent dual-core ARM making a 100 MHz square wave, lines of
    c setting a port bit up and down.

    Now, please excuse my ignorance here [1], is that more or less expensive
    than a Yamaha OPL chip? I guess they aren't made anymore, since I can't
    find anything on digikey quickly with the keywords "YMF262." So I just
    don't know how expensive they were.

    Of course, I'm sure your 75-cent dual-core ARM chips is doing something
    else too, but I got curious, and couldn't find a price.


    Best wishes, and happy electronics!

    [1] And hopefully just enjoy it.
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Wed Nov 11 02:24:16 2026
    From Newsgroup: sci.electronics.design

    On 11/10/2026 11:42 PM, john larkin wrote:
    On Tue, 10 Nov 2026 18:38:47 +0800, Johann 'Myrkraverk' Oskarsson <johann@myrkraverk.invalid> wrote:

    On 11/10/2026 6:12 PM, Liz Tuddenham wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:

    [...]
    in areas where you are ignorant you
    can end up looking pretty stupid to people who are less ignorant.

    Conversely, in areas where you are knowledgeable, you can look stupid to >>> people who are ignorant.


    And this happens all the time. It's kind of better to be a /midwit/
    than a genius [1]. I've seen that a lot, in online /special interest/
    groups, and be subject to such abuse myself. The problem when people
    get abusive, and say someone is /incompetent/, is that they are gener-
    ally wrong. First, the moron can learn, and become greater that the
    person throwing the abuse. This is an effect of time. Second, the ab-
    user rarely follows up with /why/ said abuse was thrown, so the receiver
    is left to ponder -- is there something I don't know, or are you just a
    moron?

    Now, people been around the block a few times know that today's ignora-
    mus can be a /jedi master/ in a few months, or years, and doesn't have
    to tell the rest of the group about their knowledge accumulation. So
    they always end up as /just another moron/ once the person knows better.

    This is called /lack of social skills/, and has real consequences.
    Something some members of sci.electronics.design have yet to learn.


    Best wishes, and happy jedi mastery!


    [1] If we use that term, the meme with the IQ curve I'm referring to has
    a /moron/ on the left, and a /jedi master/ on the right. The terms do
    not matter so much as long as we're being understood.

    Some people have extreme talents. It might be math, music, tennis,
    whatever. Some part of their brain is six-sigma good, often at the
    expense of other parts.

    The specialized-genius bit can also make them obnoxious.


    And som people use their /specialized-genius/ as an excuse to be ob-
    noxius. It's important to recognize those people. Sometimes you want
    to work with them, and suffer the consequences, but most of the time,
    they are best left alone.


    Best wishes, and happy social skills!
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Wed Nov 11 03:40:52 2026
    From Newsgroup: sci.electronics.design

    On 11/11/2026 12:16 AM, C||il|!n Niocl|is|!n Glost|-ir wrote:
    Johann 'Myrkraverk' Oskarsson <johann@Myrkraverk.invalid> skrev: |-----------------------------------------------------------------------| |"[. . .] Part of the | |theme of the book is the incompetence on the part of the police team --| |this includes Mike himself -- in planning and executing the undercover | |stint. [. . .] |
    |[. . .]" | |-----------------------------------------------------------------------|

    Hej!

    Do you know if there are any similar books about Scandinavian or
    Icelandic incompetent police forces or other incompetent authorities
    (e.g. courts)? (I do not need such books to be in English - I can read Swedish etc.) I asked a friend in Sweden is "Skiftet" on Netflix a
    true story. He did not know. (It is not a true story, but it is
    supposed to be realistic.) Thanks.

    Not off the top of my head, no.


    |-----------------------------------------------------------------------| |"As for how he survived, it's best you just read the book to find out. |
    | |
    |[. . .]" | |-----------------------------------------------------------------------|

    The blurb is not convincing that a person could really survive being
    shot in his head. (Trump does survive, because an ear got shot.)
    (S. HTTP://Gloucester.Insomnia247.NL/ fuer Kontaktdaten!)

    It's been a while since I read the book, so I don't remember that de-
    tail.

    It really was just a spur of the moment tsundoku, I had no idea this
    book was autobiographical when I bought it. And as far as I've been
    able to check out online, it rings true. Especially the part where
    the mobsters seemed better organized than the police force, but I'll
    refrain from spoilers.
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From john larkin@jl@htigct.com to sci.electronics.design on Tue Nov 10 12:41:02 2026
    From Newsgroup: sci.electronics.design

    On Wed, 11 Nov 2026 02:21:09 +0800, Johann 'Myrkraverk' Oskarsson <johann@myrkraverk.invalid> wrote:

    On 11/10/2026 11:59 PM, john larkin wrote:
    On Tue, 10 Nov 2026 18:20:58 +0800, Johann 'Myrkraverk' Oskarsson
    <johann@myrkraverk.invalid> wrote:

    On 11/10/2026 1:49 PM, Bill Sloman wrote:
    On 10/11/2026 2:12 am, john larkin wrote:
    On Mon, 9 Nov 2026 16:42:08 +1100, Bill Sloman <bill.sloman@ieee.org> >>>>> wrote:

    On 9/11/2026 6:27 am, Waldek Hebisch wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:
    On 7/11/2026 3:57 pm, john larkin wrote:
    On Sat, 7 Nov 2026 15:38:13 +1100, Bill Sloman <bill.sloman@ieee.org> >>>>>>>>> wrote:

    On 7/11/2026 5:54 am, john larkin wrote:
    On Fri, 6 Nov 2026 10:24:07 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and data. >>>>>>>>>>>>>>> Signed drivers

    Not a Lisp fan then?

    People love to invent weird computer languages. There are over >>>>>>>>>>>>> 3000
    and they keep coming.

    Real men code in binary using the front panel switches. >>>>>>>>>>>>
    I actually used a PDP-7 where you really did have to bootstrap >>>>>>>>>>>> it that
    way to read in a paper tape. It was a fun machine as it had a >>>>>>>>>>>> vector
    graphics display so was very good for asteroids, but cheated at >>>>>>>>>>>> chess.

    I did a lot of stuff in PDP-11 and 68K assembly languages. With >>>>>>>>>>> a few
    macros, they are almost higher-level languages.

    All high level languages get compiled into assembly language >>>>>>>>>> before they
    can be run on a computer.

    Well, no.

    Counter-example?

    Being a compiler is tedious.

    Unlike some architectures that I might name.

    A language isn't an architecture. It has to be processed into >>>>>>>>>> something
    that can be run on hardware (which will have an architecture - it is >>>>>>>>>> constructed from components that have been connected by some kind of >>>>>>>>>> architect).

    If that post were compiled to assembly language, it would be a nop. >>>>>>>>
    As a compiler, you do seem to be sub-optimal, if not downright >>>>>>>> afunctional. Ignoring implications you don't like isn't an option >>>>>>>> for a
    compiler. There can be functions they can't parse - in your case don't >>>>>>>> want to parse - but that doesn't generate a nop.

    You clearly go into an area where you are not an expert.

    I'm not pretending to be an expert, just less of a clueless newbie in >>>>>> the area that John Larkin has shown himself to be.

    I've written several compilers and several RTOS's and a bunch of
    system simulators and hundreds of application programs, a good chunk >>>>> of a million lines of code. From PDP8 to ARM and one CPU that I
    designed myself.

    And I'm sure that you are perfectly happy with all of them. Egomaniacs >>>> do go around describing themself as "stable geniuses".
    What's your definition of newbie?

    Somebody who doesn't know enough about what he his doing to be aware of >>>> the areas he doesn't know enough about. It's not so much about what you >>>> know, as what you know about the limits of your knowledge.


    And then we have the military term "unknown unknown." For instance in
    business, you cannot prevent someone from inventing something else, that >>> completely obsoletes your business.

    In /tech/, this is not an unknown unknown, because it's been done so
    many times in living memory. It can still happen in other fields, com-
    pletely blind-sighting the billionaires in charge of it.

    Some people refer to this as a /black swan/ event, which I believe comes >>>from the field of economics.

    And, to combine this topic with another one, since we were talking about >>> compilers, in the 90s, prior to the publication of GCC 2.95 [1], it
    would have been considered an /unknown unknown/ that compilers could be- >>> come malicious to security sensitive programming. I bring you [2] as
    exhibit A, and give the relevant quote.

    The short and long is that, today, the only way to access a constant
    time implementation from Rust is to write said implementation in
    assembly.

    I'll add that this is about comparing security sensitive strings, us-
    ually of fixed length [3], and /not/ an O( 1 ) algorithm.

    You can read this poster's point about why LLVM is not a polite com-
    piler at [2].

    Now, if you want to argue this conclusion, you'll have to make an acc-
    ount on Stack Overflow, and comment there.

    In other words, the only way to write constant time cryptography, is ba- >>> sically to do it in assembly, if any of GCC, LLVM, or MSVC are to com-
    pile it.

    In that sense, the /social contract/ between security sensitive program- >>> ming, and compiler makers has been broken.

    And since we are in sci.electronics.design, we can also add that hard-
    ware manufacturers typically do not assert in their manuals that a spe-
    cific instruction is constant time, so now we get to wait and see if any >>> of the x86, or ARM, or maybe even RISC-V manufacturers bring back the
    evil that was -- to my recollection -- the original 68000 shift instruc- >>> tion which spent more clock time for larger shift values.

    And to finish this off with a warning, I have read the above for adults, >>> so if you less-than-jr-high-school reading comprehension isn't up to the >>> task of figuring out what I'm saying, that's squarely on you.



    Best wishes, and happy constant time algorithms!


    [1] Given Anton Ertl's research.

    [2]
    https://stackoverflow.com/questions/44691363/how-to-compare-strings-in-constant-time

    [3] Hashed passwords are always the same length, but if the length
    changes, comparing them in constant time doesn't work; this should be
    obvious.

    I'm always impressed by how few programmers have even a rough idea of
    how much time it takes to execute their code. Guessing off by 10:1 is
    common.

    I get a scope and measure it.

    We have a 75-cent dual-core ARM making a 100 MHz square wave, lines of
    c setting a port bit up and down.

    Now, please excuse my ignorance here [1], is that more or less expensive
    than a Yamaha OPL chip? I guess they aren't made anymore, since I can't
    find anything on digikey quickly with the keywords "YMF262." So I just
    don't know how expensive they were.

    Of course, I'm sure your 75-cent dual-core ARM chips is doing something
    else too, but I got curious, and couldn't find a price.


    It's an RP2040 Raspberry Pi chip. The Pi folks recently authorized
    clocking the CPUs at 200 MHz.

    It's a wonderful part. They did everything right. We call the two CPUs
    Alice and Bob.

    Some lunatic ran a RP2040 at 1 GHz in liquid nitrogen.


    Best wishes, and happy electronics!

    [1] And hopefully just enjoy it.
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From =?UTF-8?B?Q8OzaWzDrW4=?= =?UTF-8?B?IE5pb2Nsw6Fzw61u?= =?UTF-8?B?IEdsb3N0w6lpcg==?=@thanks-to@Taf.com to sci.electronics.design on Tue Nov 10 21:18:55 2026
    From Newsgroup: sci.electronics.design

    2 of my cell phones are stolen. I stopped using a cell
    phone. Unfortunately many suppliers of things make being a client
    without a cell phone difficult.
    (S. HTTP://Gloucester.Insomnia247.NL/ fuer Kontaktdaten!)
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From ehsjr@ehsjr@verizon.net to sci.electronics.design on Tue Nov 10 16:27:18 2026
    From Newsgroup: sci.electronics.design

    On 11/10/2026 5:12 AM, Liz Tuddenham wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:

    [...]
    in areas where you are ignorant you
    can end up looking pretty stupid to people who are less ignorant.

    Conversely, in areas where you are knowledgeable, you can look stupid to people who are ignorant.


    And some who are knowledgeable use that knowledge stupidly,
    and can look stupid to most everyone, ignorant or not.
    A closed mouth gathers no foot.

    Ed
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Wed Nov 11 06:09:54 2026
    From Newsgroup: sci.electronics.design

    On 11/11/2026 5:27 AM, ehsjr wrote:
    On 11/10/2026 5:12 AM, Liz Tuddenham wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:

    [...]
    in areas where you are ignorant you
    can end up looking pretty stupid to people who are less ignorant.

    Conversely, in areas where you are knowledgeable, you can look stupid to
    people who are ignorant.


    And some who are knowledgeable use that knowledge stupidly,
    and can look stupid to most everyone, ignorant or not.
    A closed mouth gathers no foot.

    Ed

    While I am not certain that trying to /appear not stupid/ is a virtue.

    Sometimes it's quite useful to get corrections from people who know
    better. Other times, people use that /perceived stupidity/ to heap
    abuse upon the unsuspecting enquirer, with predictable results.

    In any case, trying to use /knowledge/ as a status symbol seems quite
    stupid to me; yet a lot of people behave this way.
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Wed Nov 11 06:13:48 2026
    From Newsgroup: sci.electronics.design

    On 11/11/2026 5:18 AM, C||il|!n Niocl|is|!n Glost|-ir wrote:
    2 of my cell phones are stolen. I stopped using a cell
    phone. Unfortunately many suppliers of things make being a client
    without a cell phone difficult.
    (S. HTTP://Gloucester.Insomnia247.NL/ fuer Kontaktdaten!)

    Yes, it's basically /necessary/ in life these days. You can have a cell
    phone, and just keep it in the house. It should not be stolen then. A
    tablet will do the trick too, as they -- used to -- take SIM cards, and
    can be used as a simple entertainment device. Watch movies in bed or
    whatnot.

    They can be had quite cheaply on the second hand market, even; as long
    as you're not picky.

    You can think of it as a /coping mechanism/ for a world that values peo-
    ple less and less, and technology more and more.
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From john larkin@jl@glen--canyon.com to sci.electronics.design on Tue Nov 10 18:52:17 2026
    From Newsgroup: sci.electronics.design

    On Wed, 11 Nov 2026 06:09:54 +0800, Johann 'Myrkraverk' Oskarsson <johann@myrkraverk.invalid> wrote:

    On 11/11/2026 5:27 AM, ehsjr wrote:
    On 11/10/2026 5:12 AM, Liz Tuddenham wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:

    [...]
    in areas where you are ignorant you
    can end up looking pretty stupid to people who are less ignorant.

    Conversely, in areas where you are knowledgeable, you can look stupid to >>> people who are ignorant.


    And some who are knowledgeable use that knowledge stupidly,
    and can look stupid to most everyone, ignorant or not.
    A closed mouth gathers no foot.

    Ed

    While I am not certain that trying to /appear not stupid/ is a virtue.

    Sometimes it's quite useful to get corrections from people who know
    better. Other times, people use that /perceived stupidity/ to heap
    abuse upon the unsuspecting enquirer, with predictable results.

    In any case, trying to use /knowledge/ as a status symbol seems quite
    stupid to me; yet a lot of people behave this way.

    If you are always right, you get all the blame when things go wrong.


    John Larkin
    Highland Tech Glen Canyon Design Center
    Lunatic Fringe Electronics
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Wed Nov 11 14:15:56 2026
    From Newsgroup: sci.electronics.design

    On 11/11/2026 2:37 am, john larkin wrote:
    On Tue, 10 Nov 2026 16:18:38 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 10/11/2026 2:23 pm, john larkin wrote:
    On Mon, 9 Nov 2026 23:31:54 +0800, Johann 'Myrkraverk' Oskarsson
    <johann@myrkraverk.invalid> wrote:

    On 10/31/2026 11:17 PM, Bill Sloman wrote:

    I don't make a habit of it, but there's no diplomatic way of getting the >>>>> message across, when people are ignorant (which happens quite a lot), or >>>>> stupid (which is rarer).

    Of course, you never see the same fault in yourself, you /fucking
    moron/!

    It's not anyone's fault that they are stupid. 50% of the population is
    below the median IQ.

    I'm not. It's hard to get a Ph.D. if you can't score higher than 115 on
    a IQ test, and I didn't have any trouble doing that, though I did take
    a while to finish the project. The IQ test is a cheap, simple and none
    too reliable measure of intelligence, but stupid people do pay more
    attention to IQ scores than they should.

    Be kind to the people who weren't lucky to be born beautiful geniuses
    like us.

    John Larkin isn't a genius - neither is Donald Trump - but egomaniacs do
    like to make those kinds of claims

    It is people's fault if they are nasty.

    In some situations being nasty is the only way to get the message across.

    It's the stupid, nasty, lazy, least effective way.

    But it can save a lot of time. Any nothing else works on people who have
    inflated ideas about their own competence

    Especially when you're not right.

    And you'd know a lot about that if you had any kind of realistic idea
    about your own competence. You are "usually adequate" rather than
    "reliably impressive".

    I think that's recursive.

    Egomaniacs do comfort themselves with a lot of silly ideas.
    --
    Bill Sloman, Sydney

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Wed Nov 11 14:19:25 2026
    From Newsgroup: sci.electronics.design

    On 11/11/2026 2:37 am, john larkin wrote:
    On Tue, 10 Nov 2026 16:18:38 +1100, Bill Sloman <bill.sloman@ieee.org> wrote:

    On 10/11/2026 2:23 pm, john larkin wrote:
    On Mon, 9 Nov 2026 23:31:54 +0800, Johann 'Myrkraverk' Oskarsson
    <johann@myrkraverk.invalid> wrote:

    On 10/31/2026 11:17 PM, Bill Sloman wrote:

    I don't make a habit of it, but there's no diplomatic way of
    getting the
    message across, when people are ignorant (which happens quite a
    lot), or
    stupid (which is rarer).

    Of course, you never see the same fault in yourself, you /fucking
    moron/!

    It's not anyone's fault that they are stupid. 50% of the population is
    below the median IQ.

    I'm not. It's hard to get a Ph.D. if you can't score higher than 115 on
    a IQ test, and I didn't have any trouble doing that, though I did take
    a while to finish the project. The IQ test is a cheap, simple and none
    too reliable measure of intelligence, but stupid people do pay more
    attention to IQ scores than they should.

    Be kind to the people who weren't lucky to be born beautiful geniuses
    like us.

    John Larkin isn't a genius - neither is Donald Trump - but egomaniacs do
    like to make those kinds of claims

    It is people's fault if they are nasty.

    In some situations being nasty is the only way to get the message
    across.

    It's the stupid, nasty, lazy, least effective way.

    But it can save a lot of time. And nothing else works on people who have
    inflated ideas about their own competence

    Especially when you're not right.

    And you'd know a lot about that if you had any kind of realistic idea
    about your own competence. You are "usually adequate" rather than
    "reliably impressive".

    I think that's recursive.

    Egomaniacs do comfort themselves with a lot of silly ideas.
    --
    Bill Sloman, Sydney
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From ehsjr@ehsjr@verizon.net to sci.electronics.design on Tue Nov 10 22:27:19 2026
    From Newsgroup: sci.electronics.design

    On 11/10/2026 5:09 PM, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/11/2026 5:27 AM, ehsjr wrote:
    On 11/10/2026 5:12 AM, Liz Tuddenham wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:

    [...]
    in areas where you are ignorant you
    can end up looking pretty stupid to people who are less ignorant.

    Conversely, in areas where you are knowledgeable, you can look stupid to >>> people who are ignorant.


    And some who are knowledgeable use that knowledge stupidly,
    and can look stupid to most everyone, ignorant or not.
    A closed mouth gathers no foot.

    Ed

    While I am not certain that trying to /appear not stupid/ is a virtue.

    Sometimes it's quite useful to get corrections from people who know
    better.-a Other times, people use that /perceived stupidity/ to heap
    abuse upon the unsuspecting enquirer, with predictable results.

    How pathetic to use one's (supposedly superior) intellect to heap
    abuse on the unsuspecting enquirer. It besmirches the one doing
    the heaping. What a waste of time and effort.

    Instead, the knowledgeable person could choose to express helpful
    ideas. For example, I like reading posts from Liz. She is highly knowledgeable; I've never seen her heap abuse. Rather, what she
    has to say can be helpful to those interested in the area she's
    posting about.

    Ed



    In any case, trying to use /knowledge/ as a status symbol seems quite
    stupid to me; yet a lot of people behave this way.

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Wed Nov 11 14:43:44 2026
    From Newsgroup: sci.electronics.design

    On 11/11/2026 12:40 am, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 9:08 PM, Bill Sloman wrote:
    On 10/11/2026 9:38 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 6:12 PM, Liz Tuddenham wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:

    [1] If we use that term, the meme with the IQ curve I'm referring to has >>> a /moron/ on the left, and a /jedi master/ on the right.-a The terms do
    not matter so much as long as we're being understood.

    What's obvious that you understand IQ curves to be scoring a single-
    valued parameter.

    It's well known that there are lots of components to the faculty that
    is usually described as intelligence. Quite a lot of work was done
    early on om multifactor intelligence tests, but it didn't go anywhere
    useful.

    The people doing the work didn't have much of a clue what the
    different factors might be, and didn't find any practical way of
    teasing out what they might be. When people in recent years started
    using magnetic resonance images of the brain to work out what was
    going on in the brain, where, and in what sequence, they didn't do any
    better.

    It's complicated and we need tools with a lot better discrimination
    than we have at the moment.

    If you think that you can divide the world into a linear array of
    people running from moron to the left to jedi master on the right,
    you've put yourself into the moron category.

    My late wife was a psycholinguist who spoke four languages fluently
    and was a foreign member of the US Academy of Science and a Fellow of
    the Royal Society of London, but she didn't know much about
    electronics. At least she had enough sense not to post here.


    Well, I never met your late wife, but I can tell you that I don't
    believe anything about the current "understanding" of IQ.-a It's only
    useful in one context, and that's measuring learning disability.

    It may be useful in detecting learning disability, but there are lots of different kinds of learning disabilities and you need lots of different
    tests to find out what is messing up any particular individual.

    Anything else is just handwaving and useless twiddle twaddle among the academi-
    cians.

    Through my wife I did have some contact with academics who worked on IQ testing. They weren't all that impressive, but they were streets ahead
    of handwaving and useless twiddle twaddle.

    This belief is based on actual /real life/ experience, something you
    have never experienced yourself, demonstrating time and again that you
    don't know how life works.

    I've got a least one regularly cited scientific paper

    "A microcontroller-based driver tostabilize the temperature of an
    optical stage to within 1 mK in the range 4rCo38 C, using a Peltier heat pumpand a thermistor sensor" A W Sloman, Paul Buggs, James Molloy and
    Douglas Stewart, Meas. Sci. Technol. 7 (1996) 1653rCo1664

    That does record a chunk of actual life experience. I wrote it up
    because my wife got a much better job in the Netherlands after I'd done
    the work in Cambridge and I didn't have enough to do in the first year
    that I was working in the Netherlands. The people I'd worked with were
    happy that I'd written up the work.

    And, no; I'm not going to bother explaining anything to you, because
    this is outside the scope of sci.electronics.design, and the rest of
    the people aren't asking.-a /You/ personally are not worth explaining anything to.-a You know that, don't you?

    You really haven't got a clue.

    Best wishes, and be even more miserable than usual!

    You really haven't got a clue.
    --
    Bill Sloman, Sydney

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Wed Nov 11 14:54:11 2026
    From Newsgroup: sci.electronics.design

    On 11/11/2026 2:42 am, john larkin wrote:
    On Tue, 10 Nov 2026 18:38:47 +0800, Johann 'Myrkraverk' Oskarsson <johann@myrkraverk.invalid> wrote:

    On 11/10/2026 6:12 PM, Liz Tuddenham wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:

    [...]
    in areas where you are ignorant you
    can end up looking pretty stupid to people who are less ignorant.

    Conversely, in areas where you are knowledgeable, you can look stupid to >>> people who are ignorant.


    And this happens all the time. It's kind of better to be a /midwit/
    than a genius [1]. I've seen that a lot, in online /special interest/
    groups, and be subject to such abuse myself. The problem when people
    get abusive, and say someone is /incompetent/, is that they are gener-
    ally wrong. First, the moron can learn, and become greater that the
    person throwing the abuse. This is an effect of time. Second, the ab-
    user rarely follows up with /why/ said abuse was thrown, so the receiver
    is left to ponder -- is there something I don't know, or are you just a
    moron?

    Now, people been around the block a few times know that today's ignora-
    mus can be a /jedi master/ in a few months, or years, and doesn't have
    to tell the rest of the group about their knowledge accumulation. So
    they always end up as /just another moron/ once the person knows better.

    This is called /lack of social skills/, and has real consequences.
    Something some members of sci.electronics.design have yet to learn.


    Best wishes, and happy jedi mastery!


    [1] If we use that term, the meme with the IQ curve I'm referring to has
    a /moron/ on the left, and a /jedi master/ on the right. The terms do
    not matter so much as long as we're being understood.

    Some people have extreme talents. It might be math, music, tennis,
    whatever. Some part of their brain is six-sigma good, often at the
    expense of other parts.

    The saving grace of IQ testing is that if you good in one area you tend
    to be good in others - not perhaps as exceptionally good, but still
    better than average. It may be that what is being tested for is more the absence of actual brain damage rather than some kind of superior
    processing.
    The specialized-genius bit can also make them obnoxious.

    Not in my experience.
    --
    Bill Sloman, Sydney

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Wed Nov 11 14:59:07 2026
    From Newsgroup: sci.electronics.design

    On 11/11/2026 1:52 pm, john larkin wrote:
    On Wed, 11 Nov 2026 06:09:54 +0800, Johann 'Myrkraverk' Oskarsson <johann@myrkraverk.invalid> wrote:

    On 11/11/2026 5:27 AM, ehsjr wrote:
    On 11/10/2026 5:12 AM, Liz Tuddenham wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:

    [...]
    in areas where you are ignorant you
    can end up looking pretty stupid to people who are less ignorant.

    Conversely, in areas where you are knowledgeable, you can look stupid to >>>> people who are ignorant.


    And some who are knowledgeable use that knowledge stupidly,
    and can look stupid to most everyone, ignorant or not.
    A closed mouth gathers no foot.

    Ed

    While I am not certain that trying to /appear not stupid/ is a virtue.

    Sometimes it's quite useful to get corrections from people who know
    better. Other times, people use that /perceived stupidity/ to heap
    abuse upon the unsuspecting enquirer, with predictable results.

    In any case, trying to use /knowledge/ as a status symbol seems quite
    stupid to me; yet a lot of people behave this way.

    If you are always right, you get all the blame when things go wrong.

    Not at an observation based on observed reality.

    People who are silly enough to think that they are always right don't
    get taken seriously enough to attract much blame.
    --
    Bill Sloman, Sydney

    --
    Bill Sloman, Sydney
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Wed Nov 11 15:21:19 2026
    From Newsgroup: sci.electronics.design

    On 11/11/2026 12:30 am, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 8:46 PM, Bill Sloman wrote:
    On 10/11/2026 9:22 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 6:20 PM, Johann 'Myrkraverk' Oskarsson wrote:

    And to finish this off with a warning, I have read the above for
    adults,
    so if you less-than-jr-high-school reading comprehension isn't up to
    the
    task of figuring out what I'm saying, that's squarely on you.
    Sorry, *written; I send before proof reading!

    "Sent" before proof reading. You imagine that you wrote your
    pretentious twaddle for adults, not realising that the "adults" have
    noticed that you don't have anything useful to say, and have mostly
    got around to plonking you.


    Speaking for yourself, are you?-a Because here in sci.electronics.design, you're the only one plonking.-a And you're plonking everyone else.

    "Plonking" is not reading messages from from the people plonked, and - obviously - not responding to them.

    So you're the bully who has never faced the consequences of his own
    actions.

    I rarely plonk anybody. Some people I don't bother reading - if they do
    post anything interesting other people will respond to it.

    And I'm here now, and I'll take none of it silently.-a You are a fucking bully, and you don't know it.

    Far from it.

    Now you do.

    You want to tell other people that I'm a bully. I've been posting here
    for thirty years now, so this would be well known if it were true, so it
    looks to me as if you are deceiving yourself.

    And everyone else here in sci.electronics.design knew it
    too but was too polite to tell you -- maybe.

    Jim Thompson was rude enough to claim that he'd reported me to the FBI
    for anti-American attitudes, which was a bit silly of him since I been security cleared to a level that had got me into US Army ECOM in 1970
    (back when it was at Fort Monmouth, New Jersey).

    I didn't read the 30 year backlog.

    And wouldn't have understood it if you had.

    Best wishes, and keep being miserable for the next fifty years -- at
    least!

    My statistical expectation of survival is rather shorter than that. I'd
    have to become ten years older the longest surviving human ever has so
    far to get there.
    --
    Bill Sloman, Sydney

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Wed Nov 11 15:26:44 2026
    From Newsgroup: sci.electronics.design

    On 11/11/2026 7:41 am, john larkin wrote:
    On Wed, 11 Nov 2026 02:21:09 +0800, Johann 'Myrkraverk' Oskarsson <johann@myrkraverk.invalid> wrote:

    On 11/10/2026 11:59 PM, john larkin wrote:
    On Tue, 10 Nov 2026 18:20:58 +0800, Johann 'Myrkraverk' Oskarsson
    <johann@myrkraverk.invalid> wrote:

    On 11/10/2026 1:49 PM, Bill Sloman wrote:
    On 10/11/2026 2:12 am, john larkin wrote:
    On Mon, 9 Nov 2026 16:42:08 +1100, Bill Sloman <bill.sloman@ieee.org> >>>>>> wrote:

    On 9/11/2026 6:27 am, Waldek Hebisch wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:
    On 7/11/2026 3:57 pm, john larkin wrote:
    On Sat, 7 Nov 2026 15:38:13 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 7/11/2026 5:54 am, john larkin wrote:
    On Fri, 6 Nov 2026 10:24:07 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown >>>>>>>>>>>>>>> <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and data. >>>>>>>>>>>>>>>> Signed drivers

    Not a Lisp fan then?

    People love to invent weird computer languages. There are over >>>>>>>>>>>>>> 3000
    and they keep coming.

    Real men code in binary using the front panel switches. >>>>>>>>>>>>>
    I actually used a PDP-7 where you really did have to bootstrap >>>>>>>>>>>>> it that
    way to read in a paper tape. It was a fun machine as it had a >>>>>>>>>>>>> vector
    graphics display so was very good for asteroids, but cheated at >>>>>>>>>>>>> chess.

    I did a lot of stuff in PDP-11 and 68K assembly languages. With >>>>>>>>>>>> a few
    macros, they are almost higher-level languages.

    All high level languages get compiled into assembly language >>>>>>>>>>> before they
    can be run on a computer.

    Well, no.

    Counter-example?

    Being a compiler is tedious.

    Unlike some architectures that I might name.

    A language isn't an architecture. It has to be processed into >>>>>>>>>>> something
    that can be run on hardware (which will have an architecture - it is
    constructed from components that have been connected by some kind of
    architect).

    If that post were compiled to assembly language, it would be a nop. >>>>>>>>>
    As a compiler, you do seem to be sub-optimal, if not downright >>>>>>>>> afunctional. Ignoring implications you don't like isn't an option >>>>>>>>> for a
    compiler. There can be functions they can't parse - in your case don't
    want to parse - but that doesn't generate a nop.

    You clearly go into an area where you are not an expert.

    I'm not pretending to be an expert, just less of a clueless newbie in >>>>>>> the area that John Larkin has shown himself to be.

    I've written several compilers and several RTOS's and a bunch of
    system simulators and hundreds of application programs, a good chunk >>>>>> of a million lines of code. From PDP8 to ARM and one CPU that I
    designed myself.

    And I'm sure that you are perfectly happy with all of them. Egomaniacs >>>>> do go around describing themself as "stable geniuses".
    What's your definition of newbie?

    Somebody who doesn't know enough about what he his doing to be aware of >>>>> the areas he doesn't know enough about. It's not so much about what you >>>>> know, as what you know about the limits of your knowledge.


    And then we have the military term "unknown unknown." For instance in >>>> business, you cannot prevent someone from inventing something else, that >>>> completely obsoletes your business.

    In /tech/, this is not an unknown unknown, because it's been done so
    many times in living memory. It can still happen in other fields, com- >>>> pletely blind-sighting the billionaires in charge of it.

    Some people refer to this as a /black swan/ event, which I believe comes >>> >from the field of economics.

    And, to combine this topic with another one, since we were talking about >>>> compilers, in the 90s, prior to the publication of GCC 2.95 [1], it
    would have been considered an /unknown unknown/ that compilers could be- >>>> come malicious to security sensitive programming. I bring you [2] as
    exhibit A, and give the relevant quote.

    The short and long is that, today, the only way to access a constant >>>> time implementation from Rust is to write said implementation in
    assembly.

    I'll add that this is about comparing security sensitive strings, us-
    ually of fixed length [3], and /not/ an O( 1 ) algorithm.

    You can read this poster's point about why LLVM is not a polite com-
    piler at [2].

    Now, if you want to argue this conclusion, you'll have to make an acc- >>>> ount on Stack Overflow, and comment there.

    In other words, the only way to write constant time cryptography, is ba- >>>> sically to do it in assembly, if any of GCC, LLVM, or MSVC are to com- >>>> pile it.

    In that sense, the /social contract/ between security sensitive program- >>>> ming, and compiler makers has been broken.

    And since we are in sci.electronics.design, we can also add that hard- >>>> ware manufacturers typically do not assert in their manuals that a spe- >>>> cific instruction is constant time, so now we get to wait and see if any >>>> of the x86, or ARM, or maybe even RISC-V manufacturers bring back the
    evil that was -- to my recollection -- the original 68000 shift instruc- >>>> tion which spent more clock time for larger shift values.

    And to finish this off with a warning, I have read the above for adults, >>>> so if you less-than-jr-high-school reading comprehension isn't up to the >>>> task of figuring out what I'm saying, that's squarely on you.



    Best wishes, and happy constant time algorithms!


    [1] Given Anton Ertl's research.

    [2]
    https://stackoverflow.com/questions/44691363/how-to-compare-strings-in-constant-time

    [3] Hashed passwords are always the same length, but if the length
    changes, comparing them in constant time doesn't work; this should be
    obvious.

    I'm always impressed by how few programmers have even a rough idea of
    how much time it takes to execute their code. Guessing off by 10:1 is
    common.

    I get a scope and measure it.

    We have a 75-cent dual-core ARM making a 100 MHz square wave, lines of
    c setting a port bit up and down.

    Now, please excuse my ignorance here [1], is that more or less expensive
    than a Yamaha OPL chip? I guess they aren't made anymore, since I can't
    find anything on digikey quickly with the keywords "YMF262." So I just
    don't know how expensive they were.

    Of course, I'm sure your 75-cent dual-core ARM chips is doing something
    else too, but I got curious, and couldn't find a price.


    It's an RP2040 Raspberry Pi chip. The Pi folks recently authorized
    clocking the CPUs at 200 MHz.

    It's a wonderful part. They did everything right. We call the two CPUs
    Alice and Bob.

    Some lunatic ran a RP2040 at 1 GHz in liquid nitrogen.

    What's lunatic about that? Liquid nitrogen is cheap and freely
    available. 1GHz processors aren't.
    --
    Bill Sloman, Sydney

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Wed Nov 11 15:45:19 2026
    From Newsgroup: sci.electronics.design

    On 11/11/2026 2:53 am, john larkin wrote:
    On Tue, 10 Nov 2026 16:49:55 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 10/11/2026 2:12 am, john larkin wrote:
    On Mon, 9 Nov 2026 16:42:08 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 9/11/2026 6:27 am, Waldek Hebisch wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:
    On 7/11/2026 3:57 pm, john larkin wrote:
    On Sat, 7 Nov 2026 15:38:13 +1100, Bill Sloman <bill.sloman@ieee.org> >>>>>>> wrote:

    On 7/11/2026 5:54 am, john larkin wrote:
    On Fri, 6 Nov 2026 10:24:07 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:


    <snip>

    And I'm sure that you are perfectly happy with all of them. Egomaniacs
    do go around describing themself as "stable geniuses".

    I wouldn't do that!

    So you've got more sense than Donald J. Trump - not a whole lot more,
    because you think that he has "common sense".

    What's your definition of newbie?

    Somebody who doesn't know enough about what he his doing to be aware of
    the areas he doesn't know enough about. It's not so much about what you
    know, as what you know about the limits of your knowledge.

    Electronic design talent is reality-checked by what works.

    And how well it works. There's a lot of stuff around that works well
    enough to sell, but not well enough to work all that well. I've cleaned
    up a few marginal production designs in my time.
    This worked first build. One of my guys wants to change one resistor,
    so I guess I'll let him.

    https://www.dropbox.com/scl/fi/69wwpsoc0o47cidxwust7/B360_20260813_110116.jpg?rlkey=y0sxkzswmgj3wa9lvf9ihbdio&raw=1

    A picture of the board isn't all that informative. The 360 on the label suggests that it is one more of your waveform generators. Practice
    doesn't always make perfect - it tends to re-iterate the original poor choices.

    Actually, we may change the big yellow polyfuse too. Manufacturing
    wants to kink the leads for solderability, but that lifts it up and
    then it hits the top cover and then we'd have to bend it over and that
    would look goofy.

    Life sucks.
    --
    Bill Sloman, Sydney

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From john larkin@jl@glen--canyon.com to sci.electronics.design on Tue Nov 10 21:16:23 2026
    From Newsgroup: sci.electronics.design

    On Wed, 11 Nov 2026 15:45:19 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 11/11/2026 2:53 am, john larkin wrote:
    On Tue, 10 Nov 2026 16:49:55 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 10/11/2026 2:12 am, john larkin wrote:
    On Mon, 9 Nov 2026 16:42:08 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 9/11/2026 6:27 am, Waldek Hebisch wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:
    On 7/11/2026 3:57 pm, john larkin wrote:
    On Sat, 7 Nov 2026 15:38:13 +1100, Bill Sloman <bill.sloman@ieee.org> >>>>>>>> wrote:

    On 7/11/2026 5:54 am, john larkin wrote:
    On Fri, 6 Nov 2026 10:24:07 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:


    <snip>

    And I'm sure that you are perfectly happy with all of them. Egomaniacs
    do go around describing themself as "stable geniuses".

    I wouldn't do that!

    So you've got more sense than Donald J. Trump - not a whole lot more, >because you think that he has "common sense".

    What's your definition of newbie?

    Somebody who doesn't know enough about what he his doing to be aware of
    the areas he doesn't know enough about. It's not so much about what you
    know, as what you know about the limits of your knowledge.

    Electronic design talent is reality-checked by what works.

    And how well it works. There's a lot of stuff around that works well
    enough to sell, but not well enough to work all that well. I've cleaned
    up a few marginal production designs in my time.
    This worked first build. One of my guys wants to change one resistor,
    so I guess I'll let him.

    https://www.dropbox.com/scl/fi/69wwpsoc0o47cidxwust7/B360_20260813_110116.jpg?rlkey=y0sxkzswmgj3wa9lvf9ihbdio&raw=1

    A picture of the board isn't all that informative. The 360 on the label >suggests that it is one more of your waveform generators. Practice
    doesn't always make perfect - it tends to re-iterate the original poor >choices.


    It's a 4-channel tachometer. Does waveform acquisition and quadrature
    encoders too.


    John Larkin
    Highland Tech Glen Canyon Design Center
    Lunatic Fringe Electronics
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Wed Nov 11 13:32:30 2026
    From Newsgroup: sci.electronics.design

    On 11/11/2026 11:54 AM, Bill Sloman wrote:
    On 11/11/2026 2:42 am, john larkin wrote:
    On Tue, 10 Nov 2026 18:38:47 +0800, Johann 'Myrkraverk' Oskarsson
    <johann@myrkraverk.invalid> wrote:

    On 11/10/2026 6:12 PM, Liz Tuddenham wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:

    [...]
    in areas where you are ignorant you
    can end up looking pretty stupid to people who are less ignorant.

    Conversely, in areas where you are knowledgeable, you can look
    stupid to
    people who are ignorant.


    And this happens all the time.-a It's kind of better to be a /midwit/
    than a genius [1].-a I've seen that a lot, in online /special interest/
    groups, and be subject to such abuse myself.-a The problem when people
    get abusive, and say someone is /incompetent/, is that they are gener-
    ally wrong.-a First, the moron can learn, and become greater that the
    person throwing the abuse.-a This is an effect of time.-a Second, the ab- >>> user rarely follows up with /why/ said abuse was thrown, so the receiver >>> is left to ponder -- is there something I don't know, or are you just a
    moron?

    Now, people been around the block a few times know that today's ignora-
    mus can be a /jedi master/ in a few months, or years, and doesn't have
    to tell the rest of the group about their knowledge accumulation.-a So
    they always end up as /just another moron/ once the person knows better. >>>
    This is called /lack of social skills/, and has real consequences.
    Something some members of sci.electronics.design have yet to learn.


    Best wishes, and happy jedi mastery!


    [1] If we use that term, the meme with the IQ curve I'm referring to has >>> a /moron/ on the left, and a /jedi master/ on the right.-a The terms do
    not matter so much as long as we're being understood.

    Some people have extreme talents. It might be math, music, tennis,
    whatever. Some part of their brain is six-sigma good, often at the
    expense of other parts.

    The saving grace of IQ testing is that if you good in one area you tend
    to be good in others - not perhaps as exceptionally good, but still
    better than average. It may be that what is being tested for is more the absence of actual brain damage rather than some kind of superior
    processing.
    The specialized-genius bit can also make them obnoxious.

    Not in my experience.


    You are the obnoxious /specialized genius/, so of course you wouldn't
    notice it.

    The /obvious problem/ with IQ testing is that it only tests for particu-
    lar types of puzzles solved. To be clear, the experience of it. So
    people with certain life experience tend to score higher than people
    without such life experiences. This has been a contention among politi-
    cally aligned people in recent years, so I'm sure you're familiar with
    the phenomenon.

    I could write more, but there's no point. /Why bother/. You're just a stubborn thick headed obnoxious self important academician who's deter-
    mined not to learn from people with life experiences different from
    yours.


    Best wishes, and stay miserable!
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Wed Nov 11 13:35:49 2026
    From Newsgroup: sci.electronics.design

    On 11/11/2026 12:21 PM, Bill Sloman wrote:
    On 11/11/2026 12:30 am, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 8:46 PM, Bill Sloman wrote:
    On 10/11/2026 9:22 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 6:20 PM, Johann 'Myrkraverk' Oskarsson wrote:

    And to finish this off with a warning, I have read the above for
    adults,
    so if you less-than-jr-high-school reading comprehension isn't up
    to the
    task of figuring out what I'm saying, that's squarely on you.
    Sorry, *written; I send before proof reading!

    "Sent" before proof reading. You imagine that you wrote your
    pretentious twaddle for adults, not realising that the "adults" have
    noticed that you don't have anything useful to say, and have mostly
    got around to plonking you.


    Speaking for yourself, are you?-a Because here in sci.electronics.design,
    you're the only one plonking.-a And you're plonking everyone else.

    "Plonking" is not reading messages from from the people plonked, and - obviously - not responding to them.

    So you're the bully who has never faced the consequences of his own
    actions.

    I rarely plonk anybody. Some people I don't bother reading --a if they do post anything interesting other people will respond to it.

    And I'm here now, and I'll take none of it silently.-a You are a fucking
    bully, and you don't know it.

    Far from it.

    Now you do.

    You want to tell other people that I'm a bully. I've been posting here
    for thirty years now, so this would be well known if it were true, so it looks to me as if you are deceiving yourself.

    And everyone else here in sci.electronics.design knew it
    too but was too polite to tell you -- maybe.

    Jim Thompson was rude enough to claim that he'd reported me to the FBI
    for anti-American attitudes, which was a bit silly of him since I been security cleared to a level that had got me into US Army ECOM in 1970
    (back when it was at Fort Monmouth, New Jersey).

    I didn't read the 30 year backlog.

    And wouldn't have understood it if you had.

    Best wishes, and keep being miserable for the next fifty years -- at
    least!

    My statistical expectation of survival is rather shorter than that. I'd
    have to become ten years older the longest surviving human ever has so
    far to get there.


    Don't let the statistical expectation stop you. I mean, /I want you/ to
    be here fifty years from now, and still posting your miserable drivel on sci.electronics.design, to the delight of the lurkers who simply enjoy
    the consumption of popcorn and coca-cola. A good tradition among movie
    goers going back decades, if not centuries.
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Thu Nov 12 02:11:58 2026
    From Newsgroup: sci.electronics.design

    On 11/11/2026 4:32 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/11/2026 11:54 AM, Bill Sloman wrote:
    On 11/11/2026 2:42 am, john larkin wrote:
    On Tue, 10 Nov 2026 18:38:47 +0800, Johann 'Myrkraverk' Oskarsson
    <johann@myrkraverk.invalid> wrote:

    On 11/10/2026 6:12 PM, Liz Tuddenham wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:

    <snip>

    Some people have extreme talents. It might be math, music, tennis,
    whatever. Some part of their brain is six-sigma good, often at the
    expense of other parts.

    The saving grace of IQ testing is that if you good in one area you
    tend to be good in others - not perhaps as exceptionally good, but
    still better than average. It may be that what is being tested for is
    more the absence of actual brain damage rather than some kind of
    superior processing.
    The specialized-genius bit can also make them obnoxious.

    Not in my experience.

    You are the obnoxious /specialized genius/, so of course you wouldn't
    notice it.

    I'm not any kind of genius, and while I do have skills in analog
    electronic design and fast digital design, I'm more a generalist in
    those areas than any kind of specialist.

    When I worked at EMI Central Research (from 1976 to 1979) I was exposed
    to some impressively bright people, and got on fine with most of them.
    One was a total pain in the neck, and I went to some trouble to get out
    of his project, but it wasn't his genius that was the problem, just his irresponsible antics.

    The /obvious problem/ with IQ testing is that it only tests for
    particular types of puzzles solved.-a To be clear, the experience of it.
    So people with certain life experience tend to score higher than people without such life experiences.-a This has been a contention among > politically aligned people in recent years, so I'm sure you're
    familiar with
    the phenomenon.

    It's used to explain the Flynn effect.

    https://en.wikipedia.org/wiki/Flynn_effect

    IQ tests actually include a variety of puzzles designed to test a
    variety of different intellectual skills. Nobody knows enough about the various skills for it to worth trying to test them separately.

    Multifactor intelligence tests were tried for a while, but didn't offer
    any actual advantages.

    My wife an I both scored well on IQ tests, and we both went through them looking for problems which we could solves quickly, and after we'd dealt
    with them we'd go back and tackle the ones we found harder. We'd didn't
    tackle the problems in the same order. She was better with language
    based questions, and I was better with math based stuff - we were both
    pretty good, but there was enough difference in facility that is was
    worth being picky.

    I could write more, but there's no point.-a /Why bother/.-a You're just a stubborn thick headed obnoxious self important academician who's deter-
    mined not to learn from people with life experiences different from
    yours.

    The actual problem is that you don't know enough about the subject to
    write anything worth reading. And I'm not an academic - I did the Ph.D.
    but never turned it into an academic job (unlike my wife who was a
    professor when she died and a member of a number of learned academies).

    Best wishes, and stay miserable!

    Who is supposed to be miserable?
    --
    Bill Sloman, Sydney

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From =?UTF-8?B?Q8OzaWzDrW4=?= =?UTF-8?B?IE5pb2Nsw6Fzw61u?= =?UTF-8?B?IEdsb3N0w6lpcg==?=@thanks-to@Taf.com to sci.electronics.design on Wed Nov 11 18:21:07 2026
    From Newsgroup: sci.electronics.design

    Bill Sloman <bill.sloman@IEEE.org> wrote: |-------------------------------------------------|
    |"[. . .] my mobile |
    |phone (which I do keep in my hip pocket). [. . .]|
    |[. . .]" | |-------------------------------------------------|

    When and why did this stop being illegal in Australia?
    (S. HTTP://Gloucester.Insomnia247.NL/ fuer Kontaktdaten!)
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From =?UTF-8?B?Q8OzaWzDrW4=?= =?UTF-8?B?IE5pb2Nsw6Fzw61u?= =?UTF-8?B?IEdsb3N0w6lpcg==?=@thanks-to@Taf.com to sci.electronics.design on Wed Nov 11 20:55:21 2026
    From Newsgroup: sci.electronics.design

    Don Y <blockedofcourse@foo.invalid> wrote: |--------------------------------------------------------------|
    |"Perhaps we should let hardware people write code and publicly|
    |critique their failures? Or, let software people design the |
    |hardware and let the hardware folks write the code??" | |--------------------------------------------------------------|

    VHDL
    (S. HTTP://Gloucester.Insomnia247.NL/ fuer Kontaktdaten!)
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Don Y@blockedofcourse@foo.invalid to sci.electronics.design on Wed Nov 11 14:31:55 2026
    From Newsgroup: sci.electronics.design

    On 11/9/2026 2:14 PM, Waldek Hebisch wrote:
    Don Y <blockedofcourse@foo.invalid> wrote:
    On 11/8/2026 12:27 PM, Waldek Hebisch wrote:
    You clearly go into area were you are not an expert. Google a
    bit about "undefined behaviour" in C and you will learn that
    when source code makes no sense, then C compier can do
    essentially anything, which in not so rare cases means that
    compiler simply deletes what programmer wrote. Explicit nop-s

    The compiler doesn't have THAT much license. It can't decide to
    implement arbitrary operations just because you APPEAR to be
    relying on a behavior that is undefined.

    Compiler just uses rules, so writing about appearence makes no
    sense. When optimizing compier performs a lot of transformations.
    Validity of transformations may depend on conditions. When
    checking conditions compiler assumes that program does not
    contain undefined behaviour.

    And yes, C standard gives such freedom to authors of C
    compilers:

    : undefined behavior
    : behavior, upon use of a nonportable or erroneous program construct
    : or of erroneous data, for which this document imposes no requirements

    Do I need to explain to you what "no requirements" mean?
    No one (responsibly) builds things using modern C. Because
    it is not a "practical" (as in "suitable for PRACTICE") language.

    The same bit of source code written in a particular language
    should result in the same general behavior regardless of processor
    or compiler choice.

    That contract was broken by C more than 20 years ago.

    C is now an academic exercise and playground for c.l.c types.

    All of my colleagues use C89 or K&R -- because we have to be able
    to PROVE that the source code that the customer receives and
    systematically reviews corresponds with what is happening in the
    hardware.

    Any "optimizations" have to be relatively trivial and intuitive
    and visible at the preprocessor stage; the preprocessed output
    should accuraely define what the code ACTUALLY does, not what the
    developer HOPES it will do (subject to the compiler-writer's
    whims).

    Loop unrolling, function inlining, constant folding, constant
    subexpression elimination, etc. -- all things that are OBVIOUS to
    someone inspecting the source ("Yes, I understand that STRINGLENGTH+1
    is intended to account for the NUL termination in a more visible manner"
    or "OK, I see 3+3+4+1 is a good way of representing a telephone number"
    or "Yeah, I guess you could do those steps N (constant) times in a row
    instead of wrapping them in a loop that is executed N times")

    In that way, the "source" is a writing shorthand but one that still
    completely constrains the implementation.

    E.g., I use a construct:
    object=>method(arguments)
    liberally in my code. The language doesn't support it (of course).
    But, it is known to be a representational shorthand -- something to
    unclutter the code and highlight the actual operation being performed.

    There is no surprise when you look at the preprocessor output...
    for ANY instance of such an expression.

    Even the optimizing IDL compiler is obvious in the "optimizations"
    that it makes (e.g., choosing effective marshalling sequences instead
    of leaving that up to the developer)

    After all, the role of a computer language is to convey the developer's intentions to the hardware in an unambiguous manner. ADDING ambiguity
    to that process is counterproductive.

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From =?UTF-8?B?Q8OzaWzDrW4=?= =?UTF-8?B?IE5pb2Nsw6Fzw61u?= =?UTF-8?B?IEdsb3N0w6lpcg==?=@thanks-to@Taf.com to sci.electronics.design on Wed Nov 11 21:35:33 2026
    From Newsgroup: sci.electronics.design

    Don Y <blockedofcourse@foo.invalid> wrote: |------------------------------------------------------------------|
    |"[. . .] hiring someone and then ignoring their "expert | |recommendations") happens surprisingly often. |
    | |
    |[. . .]" | |------------------------------------------------------------------|

    @book{OOSC2,
    author = {Meyer, Bertrand},
    title = {{Object-Oriented Software Construction}},
    year = {1997},
    edition = {second},
    isbn = {},
    publisher = {Prentice Hall},
    address = {},
    }
    alleges similarly. (It is an Eiffel book, but Ada is better.)

    |-------------------------------------------------------------------|
    |"Another favorite is "real-time" -- particularly "hard" vs. "soft".|
    |When most folks think it means "real fast" (it doesn't)." | |-------------------------------------------------------------------|

    True.

    |------------------------------------------------------------------|
    |"Or, that |
    |"hard" is "more difficult" than "soft" (the opposite is true -- in|
    |spades [. . .] |
    |[. . .] |
    | |
    |[. . .]" | |------------------------------------------------------------------|

    How do you mean?

    |------------------------------------------------------------------|
    |"But, so much of this is ingrained that challenging it threatens |
    |most folks' notion of their own competence so they ignore learning|
    |about those things where their ignorance is so profound. |
    | |
    |[You have to learn to work AROUND such people lest they compromise|
    |any good work]" | |------------------------------------------------------------------|

    Alas I know of no satisfactory way to work around them.
    (S. HTTP://Gloucester.Insomnia247.NL/ fuer Kontaktdaten!)
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Don Y@blockedofcourse@foo.invalid to sci.electronics.design on Wed Nov 11 14:50:30 2026
    From Newsgroup: sci.electronics.design

    |------------------------------------------------------------------|
    |"Or, that |
    |"hard" is "more difficult" than "soft" (the opposite is true -- in|
    |spades [. . .] |
    |[. . .] |
    | |
    |[. . .]" | |------------------------------------------------------------------|

    How do you mean?

    If you convince yourself (usually erroneously) that ALL of your
    deadlines are hard (non-negotiable), you overprovision your
    solution, rationalizing that you MUST HAVE those resources in
    order to meet all of those deadlines.

    Even if they stand unused in many/most instances.

    [And, your "RTOS" likely does nothing to allow you to make
    those *guarantees* as it was likely designed with the simplistic
    notion that "real fast" is "real time" (ask yourself how your
    application KNOWS that it has missed a deadline -- or, is it
    up to the user to notice that it shit the bed?)]

    This is an easy design strategy -- just figure out what the
    maximum is and ensure you have it on hand. Grouse that you
    need more memory, a faster CPU, etc. Much easier than
    deciding where the real value is in a solution!

    If, instead, you assume that not all deadlines are set in stone
    (some can be handled "late", some can be *missed*, etc.), then
    you have to come up with a STRATEGY; which do I prioritize,
    how do I decide when to abandon a particular goal, etc.

    You have to THINK about the problem instead of just naively
    assuming everything is important.

    When deadlines are soft, the "value" of a task's completion
    is no longer bi-valued (value vs. no value at all) but, rather,
    some potentially complex function of lateness. Your goal,
    then, is to sort out how to shuffle resources to "maximize value"
    (across the set of tasks) instead of the all-or-nothing decision
    that HRT suggests.

    The easy way of thinking about HRT is: if you can't make the
    deadline, STOP WORKING ON THE PROBLEM (because, by your own
    problem classification, you MUST meet the deadline).

    For how many problems would you adopt such a strategy?

    If your vehicle can't stop before hitting the obstacle,
    "take your foot off the brake!" (???)

    An SRT design tends to embed more intelligence in the
    goal setting and resource management. You *shed*
    tasks for which resource expenditure provides little
    *value* at the possible expense of other MORE valuable
    tasks.

    If there are 300 incoming missiles and your defenses can only
    handle 200 of them, do you simply give up (which is how an HRT
    problem would manifest)? Or, do you rejigger your priorities
    and decide WHICH of the 300 to target -- and how many to hold
    in reserve?

    [Of course, the simpleton mindset of politicians and military is
    to be able to handle ALL -- regardless of cost or diminishing
    returns: "Yes, general, we gave you all the weapons you need
    but there is no money left to feed your troops..."]
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From john larkin@jl@htigct.com to sci.electronics.design on Wed Nov 11 16:50:50 2026
    From Newsgroup: sci.electronics.design

    On Wed, 11 Nov 2026 14:54:11 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 11/11/2026 2:42 am, john larkin wrote:
    On Tue, 10 Nov 2026 18:38:47 +0800, Johann 'Myrkraverk' Oskarsson
    <johann@myrkraverk.invalid> wrote:

    On 11/10/2026 6:12 PM, Liz Tuddenham wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:

    [...]
    in areas where you are ignorant you
    can end up looking pretty stupid to people who are less ignorant.

    Conversely, in areas where you are knowledgeable, you can look stupid to >>>> people who are ignorant.


    And this happens all the time. It's kind of better to be a /midwit/
    than a genius [1]. I've seen that a lot, in online /special interest/
    groups, and be subject to such abuse myself. The problem when people
    get abusive, and say someone is /incompetent/, is that they are gener-
    ally wrong. First, the moron can learn, and become greater that the
    person throwing the abuse. This is an effect of time. Second, the ab-
    user rarely follows up with /why/ said abuse was thrown, so the receiver >>> is left to ponder -- is there something I don't know, or are you just a
    moron?

    Now, people been around the block a few times know that today's ignora-
    mus can be a /jedi master/ in a few months, or years, and doesn't have
    to tell the rest of the group about their knowledge accumulation. So
    they always end up as /just another moron/ once the person knows better. >>>
    This is called /lack of social skills/, and has real consequences.
    Something some members of sci.electronics.design have yet to learn.


    Best wishes, and happy jedi mastery!


    [1] If we use that term, the meme with the IQ curve I'm referring to has >>> a /moron/ on the left, and a /jedi master/ on the right. The terms do
    not matter so much as long as we're being understood.

    Some people have extreme talents. It might be math, music, tennis,
    whatever. Some part of their brain is six-sigma good, often at the
    expense of other parts.

    The saving grace of IQ testing is that if you good in one area you tend
    to be good in others - not perhaps as exceptionally good, but still
    better than average. It may be that what is being tested for is more the >absence of actual brain damage rather than some kind of superior
    processing.
    The specialized-genius bit can also make them obnoxious.

    Not in my experience.

    Sometimes you actually, probably accidentally, approach humor.

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From john larkin@jl@htigct.com to sci.electronics.design on Wed Nov 11 16:53:22 2026
    From Newsgroup: sci.electronics.design

    On Wed, 11 Nov 2026 15:26:44 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 11/11/2026 7:41 am, john larkin wrote:
    On Wed, 11 Nov 2026 02:21:09 +0800, Johann 'Myrkraverk' Oskarsson
    <johann@myrkraverk.invalid> wrote:

    On 11/10/2026 11:59 PM, john larkin wrote:
    On Tue, 10 Nov 2026 18:20:58 +0800, Johann 'Myrkraverk' Oskarsson
    <johann@myrkraverk.invalid> wrote:

    On 11/10/2026 1:49 PM, Bill Sloman wrote:
    On 10/11/2026 2:12 am, john larkin wrote:
    On Mon, 9 Nov 2026 16:42:08 +1100, Bill Sloman <bill.sloman@ieee.org> >>>>>>> wrote:

    On 9/11/2026 6:27 am, Waldek Hebisch wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:
    On 7/11/2026 3:57 pm, john larkin wrote:
    On Sat, 7 Nov 2026 15:38:13 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 7/11/2026 5:54 am, john larkin wrote:
    On Fri, 6 Nov 2026 10:24:07 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown >>>>>>>>>>>>>>>> <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and data. >>>>>>>>>>>>>>>>> Signed drivers

    Not a Lisp fan then?

    People love to invent weird computer languages. There are over >>>>>>>>>>>>>>> 3000
    and they keep coming.

    Real men code in binary using the front panel switches. >>>>>>>>>>>>>>
    I actually used a PDP-7 where you really did have to bootstrap >>>>>>>>>>>>>> it that
    way to read in a paper tape. It was a fun machine as it had a >>>>>>>>>>>>>> vector
    graphics display so was very good for asteroids, but cheated at >>>>>>>>>>>>>> chess.

    I did a lot of stuff in PDP-11 and 68K assembly languages. With >>>>>>>>>>>>> a few
    macros, they are almost higher-level languages.

    All high level languages get compiled into assembly language >>>>>>>>>>>> before they
    can be run on a computer.

    Well, no.

    Counter-example?

    Being a compiler is tedious.

    Unlike some architectures that I might name.

    A language isn't an architecture. It has to be processed into >>>>>>>>>>>> something
    that can be run on hardware (which will have an architecture - it is
    constructed from components that have been connected by some kind of
    architect).

    If that post were compiled to assembly language, it would be a nop. >>>>>>>>>>
    As a compiler, you do seem to be sub-optimal, if not downright >>>>>>>>>> afunctional. Ignoring implications you don't like isn't an option >>>>>>>>>> for a
    compiler. There can be functions they can't parse - in your case don't
    want to parse - but that doesn't generate a nop.

    You clearly go into an area where you are not an expert.

    I'm not pretending to be an expert, just less of a clueless newbie in >>>>>>>> the area that John Larkin has shown himself to be.

    I've written several compilers and several RTOS's and a bunch of >>>>>>> system simulators and hundreds of application programs, a good chunk >>>>>>> of a million lines of code. From PDP8 to ARM and one CPU that I
    designed myself.

    And I'm sure that you are perfectly happy with all of them. Egomaniacs >>>>>> do go around describing themself as "stable geniuses".
    What's your definition of newbie?

    Somebody who doesn't know enough about what he his doing to be aware of >>>>>> the areas he doesn't know enough about. It's not so much about what you >>>>>> know, as what you know about the limits of your knowledge.


    And then we have the military term "unknown unknown." For instance in >>>>> business, you cannot prevent someone from inventing something else, that >>>>> completely obsoletes your business.

    In /tech/, this is not an unknown unknown, because it's been done so >>>>> many times in living memory. It can still happen in other fields, com- >>>>> pletely blind-sighting the billionaires in charge of it.

    Some people refer to this as a /black swan/ event, which I believe comes >>>> >from the field of economics.

    And, to combine this topic with another one, since we were talking about >>>>> compilers, in the 90s, prior to the publication of GCC 2.95 [1], it
    would have been considered an /unknown unknown/ that compilers could be- >>>>> come malicious to security sensitive programming. I bring you [2] as >>>>> exhibit A, and give the relevant quote.

    The short and long is that, today, the only way to access a constant >>>>> time implementation from Rust is to write said implementation in >>>>> assembly.

    I'll add that this is about comparing security sensitive strings, us- >>>>> ually of fixed length [3], and /not/ an O( 1 ) algorithm.

    You can read this poster's point about why LLVM is not a polite com- >>>>> piler at [2].

    Now, if you want to argue this conclusion, you'll have to make an acc- >>>>> ount on Stack Overflow, and comment there.

    In other words, the only way to write constant time cryptography, is ba- >>>>> sically to do it in assembly, if any of GCC, LLVM, or MSVC are to com- >>>>> pile it.

    In that sense, the /social contract/ between security sensitive program- >>>>> ming, and compiler makers has been broken.

    And since we are in sci.electronics.design, we can also add that hard- >>>>> ware manufacturers typically do not assert in their manuals that a spe- >>>>> cific instruction is constant time, so now we get to wait and see if any >>>>> of the x86, or ARM, or maybe even RISC-V manufacturers bring back the >>>>> evil that was -- to my recollection -- the original 68000 shift instruc- >>>>> tion which spent more clock time for larger shift values.

    And to finish this off with a warning, I have read the above for adults, >>>>> so if you less-than-jr-high-school reading comprehension isn't up to the >>>>> task of figuring out what I'm saying, that's squarely on you.



    Best wishes, and happy constant time algorithms!


    [1] Given Anton Ertl's research.

    [2]
    https://stackoverflow.com/questions/44691363/how-to-compare-strings-in-constant-time

    [3] Hashed passwords are always the same length, but if the length
    changes, comparing them in constant time doesn't work; this should be >>>>> obvious.

    I'm always impressed by how few programmers have even a rough idea of
    how much time it takes to execute their code. Guessing off by 10:1 is
    common.

    I get a scope and measure it.

    We have a 75-cent dual-core ARM making a 100 MHz square wave, lines of >>>> c setting a port bit up and down.

    Now, please excuse my ignorance here [1], is that more or less expensive >>> than a Yamaha OPL chip? I guess they aren't made anymore, since I can't >>> find anything on digikey quickly with the keywords "YMF262." So I just
    don't know how expensive they were.

    Of course, I'm sure your 75-cent dual-core ARM chips is doing something
    else too, but I got curious, and couldn't find a price.


    It's an RP2040 Raspberry Pi chip. The Pi folks recently authorized
    clocking the CPUs at 200 MHz.

    It's a wonderful part. They did everything right. We call the two CPUs
    Alice and Bob.

    Some lunatic ran a RP2040 at 1 GHz in liquid nitrogen.

    What's lunatic about that? Liquid nitrogen is cheap and freely
    available. 1GHz processors aren't.

    I think they also pushed the 1.2v core voltage to 5.

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Thu Nov 12 13:16:31 2026
    From Newsgroup: sci.electronics.design

    On 11/11/2026 4:35 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/11/2026 12:21 PM, Bill Sloman wrote:
    On 11/11/2026 12:30 am, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 8:46 PM, Bill Sloman wrote:
    On 10/11/2026 9:22 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 6:20 PM, Johann 'Myrkraverk' Oskarsson wrote:

    <snip>

    Best wishes, and keep being miserable for the next fifty years -- at
    least!

    My statistical expectation of survival is rather shorter than that.
    I'd have to become ten years older than longest surviving human ever
    has so far to get there.

    Don't let the statistical expectation stop you.

    Statistical expectation isn't any kind of barrier. It's just a
    prediction based on what has happened in the past.

    Modern medical science does keep people alive for longer than they used
    to manage to survive, but 133 would be a very ripe old age.


    I mean, /I want you/ to
    be here fifty years from now, and still posting your miserable drivel on sci.electronics.design, to the delight of the lurkers who simply enjoy
    the consumption of popcorn and coca-cola.-a A good tradition among movie goers going back decades, if not centuries.

    What you perceive as miserable drivel may not look quite so miserable
    and drivel-like to better-informed observers (who seem to be fairly
    numerous).

    And I really don't care what you want. The peanut gallery isn't a discriminating audience.
    --
    Bill Sloman, Sydney

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Thu Nov 12 13:21:32 2026
    From Newsgroup: sci.electronics.design

    On 12/11/2026 5:21 am, C||il|!n Niocl|is|!n Glost|-ir wrote:
    Bill Sloman <bill.sloman@IEEE.org> wrote: |-------------------------------------------------|
    |"[. . .] my mobile |
    |phone (which I do keep in my hip pocket). [. . .]|
    |[. . .]" | |-------------------------------------------------|

    When and why did this stop being illegal in Australia?
    (S. HTTP://Gloucester.Insomnia247.NL/ fuer Kontaktdaten!)

    When was it ever illegal anywhere? My earlier mobile phones weren't
    compact enough to fit, but the Samsung S23 is thin enough to make it practical.
    --
    Bill Sloman, Sydney
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Thu Nov 12 13:39:17 2026
    From Newsgroup: sci.electronics.design

    On 12/11/2026 11:50 am, john larkin wrote:
    On Wed, 11 Nov 2026 14:54:11 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 11/11/2026 2:42 am, john larkin wrote:
    On Tue, 10 Nov 2026 18:38:47 +0800, Johann 'Myrkraverk' Oskarsson
    <johann@myrkraverk.invalid> wrote:

    On 11/10/2026 6:12 PM, Liz Tuddenham wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:


    The saving grace of IQ testing is that if you good in one area you tend
    to be good in others - not perhaps as exceptionally good, but still
    better than average. It may be that what is being tested for is more the
    absence of actual brain damage rather than some kind of superior
    processing.

    The specialized-genius bit can also make them obnoxious.

    Not in my experience.

    Sometimes you actually, probably accidentally, approach humor.

    Back in about 1968 I wrote a few scripts for the Melbourne University Architects Revue. Some of them got into the revue and got laughs. Rowan Atkinson did much better when he was at Cambridge, and dropped his
    master's degree in electronics in favour of a theatrical career.

    Your own credentials as a judge of humour are even less impressive.

    There were a few specialised genius types around EMI Central Research
    when I worked there, and only one of them was obnoxious. It wasn't his
    genius that was the problem, but his irresponsible antics.

    William S. Percival was probably cleverer, but quite easy to get on with.

    https://en.wikipedia.org/wiki/Distributed_amplifier
    --
    Bill Sloman, Sydney


    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Thu Nov 12 15:35:41 2026
    From Newsgroup: sci.electronics.design

    On 10/11/2026 10:02 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 5:14 AM, Waldek Hebisch wrote:
    Actually, at '-O2' version of 'gcc' that I use replaces explicit
    division by 0 by the 'ud2' instruction.

    Shouldn't that be considered /erroneous compilation/?-a I mean, you can
    write a signal handler that catches all signals, and isn't /division by
    zero/ one of the Posix signals?-a Is the compiler prescient and makes
    sure this only happens if the binary is ever run when the user isn't ex- plictly expecting a division by zero -- rather than /invalid opcode/?

    Does the question bother you?-a Am I too high, or too low on the /moron/
    vs. /jedi master/ axis to be understood?-a Should we put all the GCC devs
    in therapy for believing they are prescient?

    I imagine that there are situations where the programmer can be
    confident that the divisor is never going to be zero.

    A cautious programmer might test for a zero divisor and run an error
    recovery routine rather than doing the division.

    To me it does look as if you've posted an under-informed response.
    --
    Bill Sloman, Sydney
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Don Y@blockedofcourse@foo.invalid to sci.electronics.design on Thu Nov 12 05:58:44 2026
    From Newsgroup: sci.electronics.design

    On 11/11/2026 2:50 PM, Don Y wrote:
    You have to THINK about the problem instead of just naively
    assuming everything is important.

    When deadlines are soft, the "value" of a task's completion
    is no longer bi-valued (value vs. no value at all) but, rather,
    some potentially complex function of lateness.-a Your goal,
    then, is to sort out how to shuffle resources to "maximize value"
    (across the set of tasks) instead of the all-or-nothing decision
    that HRT suggests.

    An SRT design tends to embed more intelligence in the
    goal setting and resource management.-a You *shed*
    tasks for which resource expenditure provides little
    *value* at the possible expense of other MORE valuable
    tasks.
    Here's a simple, practical example that illustrates the costs and
    value of a single "task" and how it might be instantiated:

    E.g., if I want to "water the citrus", I need the hardware interface to the irrigation valves that is present on the irrigation controller. Otherwise,
    I can't "gate" water to that portion of the yard. I need the "grey water" supply and its metering hardware (another node) to source the water. And, possibly, the municipal water supply and its metering hardware -- in the
    event that there is insufficient grey water available.

    I need the database that tracks weather history so I can determine how *much* water will be required (based on precipitation, temperatures, wind, etc.).
    If it rained recently -- or, if it has been hot/windy/dry -- then the current needs of those plants will differ.

    And, I need memory and "MIPS-seconds" on each of these nodes to actually do
    the "work" required. So, those devices must be powered up and have sufficient "spare" resources to host the activities that I will require of them.

    During daylight hours, I'll likely have a surplus of electrical power
    (assuming the power system's estimate of the remaining available solar
    power will suffice to top of the storage batteries). At night, power
    may be in shorter supply (or, "more costly" so this has to be justified
    by the "value" of getting the task accomplished!). OTOH, irrigation is
    more effective at night as evaporative losses are greatly reduced and
    the flora will be most receptive to water uptake as the sun rises.
    OToOH, if the power has failed, then power will be even more precious;
    perhaps risk the fruit taking a hit so some more important task can be accomplished?

    The scheduler now has to find a time when those resources are available
    (or, can be MADE available -- powering hardware on if presently off,
    evicting jobs of lower priority, etc.) and then admit the task to the
    run queue.

    *Hoping* that conditions don't change significantly before it finishes
    (a minute from now, an occupant may want to take a shower and the drop
    in water pressure caused by the huge load represented by the citrus
    might be objectionable -- so, that task will have to be paused or
    aborted)

    If I had taken the attitude that "everything is important and of immediate need", then I would have had to increase the size of the grey water tank
    (and the pump and piping that distributes its contents as I might have had
    to irrigate EVERYTHING on the property at the same instant!), the diameter
    of the municipal water main to handle those cases where the grey water
    supply was inadequate, the size of the storage battery in case lots of things wanted to happen "after dark", the number and capabilities of the CPUs
    doing the work, etc.

    [Irrigation is much easier for people to wrap their heads around than
    the effort and resources required to train a speech model or update a
    scene analysis, hence my reliance on it as an example]

    This, of course, is just stupid. A smarter approach is to juggle priorities
    so you can do everything with *less*. By recognizing that not everything
    has to happen immediately. By treating most "deadlines" as ADVISORY and
    not HARD REQUIREMENTS.

    [You can see there are far more design decisions involved in this approach!] --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Fri Nov 13 01:31:39 2026
    From Newsgroup: sci.electronics.design

    On 11/11/2026 2:27 pm, ehsjr wrote:
    On 11/10/2026 5:09 PM, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/11/2026 5:27 AM, ehsjr wrote:
    On 11/10/2026 5:12 AM, Liz Tuddenham wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:

    [...]
    in areas where you are ignorant you
    can end up looking pretty stupid to people who are less ignorant.

    Conversely, in areas where you are knowledgeable, you can look
    stupid to
    people who are ignorant.


    And some who are knowledgeable use that knowledge stupidly,
    and can look stupid to most everyone, ignorant or not.
    A closed mouth gathers no foot.

    Ed

    While I am not certain that trying to /appear not stupid/ is a virtue.

    Sometimes it's quite useful to get corrections from people who know
    better.-a Other times, people use that /perceived stupidity/ to heap
    abuse upon the unsuspecting enquirer, with predictable results.

    How pathetic to use one's (supposedly superior) intellect to heap
    abuse on the unsuspecting enquirer.-a It besmirches the one doing
    the heaping. What a waste of time and effort.

    What's being complained about here isn't so much "heaping abuse" as not providing the level of flattery that some posters feel they deserve.

    Instead, the knowledgeable person could choose to express helpful
    ideas. For example, I like reading posts from Liz.-a She is highly knowledgeable; I've never seen her heap abuse.-a Rather, what she
    has to say can be helpful to those interested in the area she's
    posting about.

    She is frequently helpful.

    In any case, trying to use /knowledge/ as a status symbol seems quite
    stupid to me; yet a lot of people behave this way.

    Correcting errors isn't "using knowledge as a status symbol", except in
    the sense that being able to correct errors does involve asserting a
    superior status, but since the alternative is to leave errors unremarked
    and uncorrected, it is also taking responsibility.
    --
    Bill Sloman, Sydney

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From antispam@antispam@fricas.org (Waldek Hebisch) to sci.electronics.design on Thu Nov 12 16:36:40 2026
    From Newsgroup: sci.electronics.design

    Don Y <blockedofcourse@foo.invalid> wrote:
    On 11/9/2026 2:14 PM, Waldek Hebisch wrote:
    Don Y <blockedofcourse@foo.invalid> wrote:
    On 11/8/2026 12:27 PM, Waldek Hebisch wrote:
    You clearly go into area were you are not an expert. Google a
    bit about "undefined behaviour" in C and you will learn that
    when source code makes no sense, then C compier can do
    essentially anything, which in not so rare cases means that
    compiler simply deletes what programmer wrote. Explicit nop-s

    The compiler doesn't have THAT much license. It can't decide to
    implement arbitrary operations just because you APPEAR to be
    relying on a behavior that is undefined.

    Compiler just uses rules, so writing about appearence makes no
    sense. When optimizing compier performs a lot of transformations.
    Validity of transformations may depend on conditions. When
    checking conditions compiler assumes that program does not
    contain undefined behaviour.

    And yes, C standard gives such freedom to authors of C
    compilers:

    : undefined behavior
    : behavior, upon use of a nonportable or erroneous program construct
    : or of erroneous data, for which this document imposes no requirements

    Do I need to explain to you what "no requirements" mean?
    No one (responsibly) builds things using modern C. Because
    it is not a "practical" (as in "suitable for PRACTICE") language.

    The "problem" first appeared in context of optimizing Fortran
    compliers and AFAIK while wording is different and specific
    conditions differs in effect Fortran standard allows the
    same. Also Pascal standard. IIUC Ada mandates runtime checks
    for a lot of things and many other things lead to compile time
    errors. But programmer determided to write programs in C
    style still can inject equivalent of undefined behaviour with
    consequences the same as in C.

    AFAIK a lot of folks use C for important software. Some claim
    that their code contains no bugs, so in particular no undefiend
    behaviour. Some use tools like CompCert. IIUC some safety
    standard certify _object_ code, it does not matter how good or
    buggy is C compiler that generated object code, you are
    supposed to review generated code and verify that it is doing
    what it is supposed to do (and presumably you need to keep
    appropriate document trail to show that you really did that).

    Some people think that software is serious when it brings money
    to its creators. AFAIK most optimizations in gcc and clang
    are funded by commercial companies which see gains from
    those optimizations. Most of this is motivated by C++,
    where you really need those optimizations for decent speed.
    But problems with undefined bahaviour are almost the same.

    The same bit of source code written in a particular language
    should result in the same general behavior regardless of processor
    or compiler choice.

    It will, if it is defined code.

    That contract was broken by C more than 20 years ago.

    C is now an academic exercise and playground for c.l.c types.

    All of my colleagues use C89 or K&R -- because we have to be able
    to PROVE that the source code that the customer receives and
    systematically reviews corresponds with what is happening in the
    hardware.

    Use of K&R C for new code looks irresponsible to me: checking that
    code accepted by K&R compiler has no bugs is more effort than
    the same for modern C. C89 looks reasonable, but it gains you
    a little. Main thing are compilers. There are two mainstream
    ones with multiple targets, that is gcc and clang. There were
    few respectable compilers from specific vendors targeting their
    targets, but most seem to be gone now. And all compilers beside
    gcc and clang seem to get much less use and consequently are
    less tested.

    When using C compiler you may have trouble for various reasons.
    One thing are ordinary bugs in C code. Other thing are bugs
    which cause undefined behaviour. Third thing are compiler bugs,
    when source code is well defined, but compiler produces wrong
    object code. AFAICS modern compilers are much more helpful
    with catching ordinary bugs. Concerning wrong code, they
    seem to be better, but larger size of compiler and increased
    complexity may leave some doubts. Undefined behaviour could
    catch you even 35 years ago: at that time users complained
    that their working code is broken by new C compilers. Users
    learned that old code was wrong and how to write correct
    code. I do not see much difference now: you learn to write
    code in a different way, which has additional advantage that
    code avoinding risky constructs usually is also clearer.

    Any "optimizations" have to be relatively trivial and intuitive
    and visible at the preprocessor stage; the preprocessed output
    should accuraely define what the code ACTUALLY does, not what the
    developer HOPES it will do (subject to the compiler-writer's
    whims).

    You preach specifications yet you have trouble that compiler also
    have specification. Stay withing specified (defined) constructs
    and your code will have expected effect. If you expect more
    than is specified, then you expect too much.

    Concerning "what the code ACTUALLY does", I expect compiler
    to delete checks that are known to pass at compile time
    (such checks may be generated from macros and I want to keep
    my macros). I expect compiler to expand inline simple functions,
    when it leads to smaller code size. In general generated
    code may look quite different than expectation of some
    folks. What counts is the effect.

    For delay loop I have assembly code: C compiler may and should
    collapse to nothing loops that take time but otherwise do not
    contribute to the result. And I find atempts to cripple
    optimization in order to preserve delay loop to be too
    unreliable. But otherwise C allows me to write what
    should be done and compiler generates appropriate code.

    Loop unrolling, function inlining, constant folding, constant
    subexpression elimination, etc. -- all things that are OBVIOUS to
    someone inspecting the source ("Yes, I understand that STRINGLENGTH+1
    is intended to account for the NUL termination in a more visible manner"
    or "OK, I see 3+3+4+1 is a good way of representing a telephone number"
    or "Yeah, I guess you could do those steps N (constant) times in a row instead of wrapping them in a loop that is executed N times")

    In that way, the "source" is a writing shorthand but one that still completely constrains the implementation.

    E.g., I use a construct:
    object=>method(arguments)
    liberally in my code. The language doesn't support it (of course).
    But, it is known to be a representational shorthand -- something to
    unclutter the code and highlight the actual operation being performed.

    There is no surprise when you look at the preprocessor output...
    for ANY instance of such an expression.

    Even the optimizing IDL compiler is obvious in the "optimizations"
    that it makes (e.g., choosing effective marshalling sequences instead
    of leaving that up to the developer)

    That is slightly different thing. Optimizations require some effort.
    If you have preprocessor/precompiler it makes no sense to duplicate optimizations done by C compiler. Only things that C compiler
    can not optimize are worth optimization in a precompiler and only
    when they are important enough. More optimizations in C compiler
    means that your precompiler may be simpler. For example
    precompiler may naively insert code trusting that C compiler
    will remove it in cases where it is not needed.
    --
    Waldek Hebisch
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From ehsjr@ehsjr@verizon.net to sci.electronics.design on Thu Nov 12 12:19:35 2026
    From Newsgroup: sci.electronics.design

    On 11/12/2026 9:31 AM, Bill Sloman wrote:
    On 11/11/2026 2:27 pm, ehsjr wrote:
    On 11/10/2026 5:09 PM, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/11/2026 5:27 AM, ehsjr wrote:
    On 11/10/2026 5:12 AM, Liz Tuddenham wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:

    [...]
    in areas where you are ignorant you
    can end up looking pretty stupid to people who are less ignorant.

    Conversely, in areas where you are knowledgeable, you can look
    stupid to
    people who are ignorant.


    And some who are knowledgeable use that knowledge stupidly,
    and can look stupid to most everyone, ignorant or not.
    A closed mouth gathers no foot.

    Ed

    While I am not certain that trying to /appear not stupid/ is a virtue.

    Sometimes it's quite useful to get corrections from people who know
    better.-a Other times, people use that /perceived stupidity/ to heap
    abuse upon the unsuspecting enquirer, with predictable results.

    How pathetic to use one's (supposedly superior) intellect to heap
    abuse on the unsuspecting enquirer.-a It besmirches the one doing
    the heaping. What a waste of time and effort.

    What's being complained about here isn't so much "heaping abuse" as not providing the level of flattery that some posters feel they deserve.

    Johann used that term "heaping abuse" - others may or may not
    consider it accurate. That said, it is best to not flatter in
    any case, whatever the poster's motive was. Flattery is not a
    good thing. Nor is abuse. Just ignore posts where you think
    the poster is looking for flattery.




    Instead, the knowledgeable person could choose to express helpful
    ideas. For example, I like reading posts from Liz.-a She is highly
    knowledgeable; I've never seen her heap abuse.-a Rather, what she
    has to say can be helpful to those interested in the area she's
    posting about.

    She is frequently helpful.

    In any case, trying to use /knowledge/ as a status symbol seems quite
    stupid to me; yet a lot of people behave this way.

    I did not post the above, Johann did. Your reply to it (below)
    belongs elsewhere.

    Ed




    Correcting errors isn't "using knowledge as a status symbol", except in
    the sense that being able to correct errors does involve asserting a superior status, but since the alternative is to leave errors unremarked
    and uncorrected, it is also taking responsibility.


    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From =?UTF-8?B?Q8OzaWzDrW4=?= =?UTF-8?B?IE5pb2Nsw6Fzw61u?= =?UTF-8?B?IEdsb3N0w6lpcg==?=@thanks-to@Taf.com to sci.electronics.design on Thu Nov 12 18:15:17 2026
    From Newsgroup: sci.electronics.design

    Thanks for this answer.
    (S. HTTP://Gloucester.Insomnia247.NL/ fuer Kontaktdaten!)
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Don Y@blockedofcourse@foo.invalid to sci.electronics.design on Thu Nov 12 13:25:06 2026
    From Newsgroup: sci.electronics.design

    On 11/12/2026 9:36 AM, Waldek Hebisch wrote:
    The compiler doesn't have THAT much license. It can't decide to
    implement arbitrary operations just because you APPEAR to be
    relying on a behavior that is undefined.

    Compiler just uses rules, so writing about appearence makes no
    sense. When optimizing compier performs a lot of transformations.
    Validity of transformations may depend on conditions. When
    checking conditions compiler assumes that program does not
    contain undefined behaviour.

    And yes, C standard gives such freedom to authors of C
    compilers:

    : undefined behavior
    : behavior, upon use of a nonportable or erroneous program construct
    : or of erroneous data, for which this document imposes no requirements

    Do I need to explain to you what "no requirements" mean?
    No one (responsibly) builds things using modern C. Because
    it is not a "practical" (as in "suitable for PRACTICE") language.

    The "problem" first appeared in context of optimizing Fortran
    compliers and AFAIK while wording is different and specific
    conditions differs in effect Fortran standard allows the
    same. Also Pascal standard. IIUC Ada mandates runtime checks
    for a lot of things and many other things lead to compile time
    errors. But programmer determided to write programs in C
    style still can inject equivalent of undefined behaviour with
    consequences the same as in C.

    C's initial appeal was its simplicity (it is a SMALL language) and
    "closeness to the hardware.

    The first normative standard was essential because it created
    THE language (instead of a collection of different interpretations
    of the language). And, codified things that should have been
    defined earlier (function prototypes, the standard library).
    It also HELPED the programmer with volatile, enum, signed,
    void, const, etc.

    But, every revision in beyond that seemed to be geared towards
    the compiler vendor; allowing more and more leeway (usually
    without any requirement to HELP the programmer).

    Given that compilers run in resource rich environments and
    tend to be (re)written fewer times than applications, this
    seems a wrong emphasis.

    And, given that the standard is free to say ANYTHING regarding
    "requirements", it could just as easily have said "When encountering
    undefined behavior, a loud klaxon should sound, accompanied by
    blinkenlites and a complete refusal to emit ANY code, thus
    drawing the programmer's attention to the "problem".

    There should be a standard switch to say "go back to the prior
    standard definition". Etc. Or, to flag any behavior that
    may have changed since the last time you compiled this codebase.
    How many people REALLY care about how long a build takes?
    Can they not afford a faster development platform to provide
    the speed they need, so the poor compiler writers don't have to
    struggle so hard to speed up *their* programs?

    Instead, the onus is put on the programmer to figure out how to
    coax the compiler to emit code for sources that HAVE BEEN WORKING FOR
    YEARS under prior compilers and language definitions.

    A language is supposed to facilitate communication (in this case, of
    ideas, concepts, constructs and mechanisms). Instead, C has continued
    to muddy the waters until folks will eventually abandon it (or, cling
    to older versions that are better behaved).

    AFAIK a lot of folks use C for important software. Some claim
    that their code contains no bugs, so in particular no undefiend
    behaviour. Some use tools like CompCert. IIUC some safety
    standard certify _object_ code, it does not matter how good or
    buggy is C compiler that generated object code, you are
    supposed to review generated code and verify that it is doing
    what it is supposed to do (and presumably you need to keep
    appropriate document trail to show that you really did that).

    Have you ever tried to validate a substantial piece of software?
    It is a grueling task and requires considerable expertise on both
    sides of the table (client and provider). Adding the *need* for
    sophisticated language lawyers just makes that harder:
    "What happens if X+Y oveflows? And, what about A+B? C-D? P*Q?"
    Make a change to the codebase and repeat the exercise (esp if the
    compiler has license to do code motion optimizations, etc.)

    The original language had a relatively intuitive mapping onto the hardware/machine language. So, could be used as a shorthand for it,
    saving all of those extra lines of ASM from having to be meticulously inspected. A modest piece of code is O(10KLoC). If you were
    reviewing the ASM associated with that, expect some multiple.

    If the compiler is free to jigger things around so they don't
    directly correspond to the C source, then good luck trying to
    understand what the high level abstractions implemented by a
    particular section of ASM might be.

    [It's trivial to disassemble a piece of code. Imagine trying to
    figure out what it *does* from that disassembly, lacking a
    direct "explanation" of each code section by the accompanying
    "source"]

    Some people think that software is serious when it brings money
    to its creators. AFAIK most optimizations in gcc and clang
    are funded by commercial companies which see gains from
    those optimizations. Most of this is motivated by C++,
    where you really need those optimizations for decent speed.
    But problems with undefined bahaviour are almost the same.

    The embedded system market generates more applications than
    the desktop worl by orders of magnitude. And, the desktop
    world operates in a resource rich environment. Along with
    readily facilitating software revision and updates. These
    things aren't available in the embedded systems world -- where
    people BUILD things and customers expect them to work.
    (You don't even consider updating the software in your
    refrigerator, microwave oven, mouse, keyboard, monitor,
    doorbell, etc. because you would view that as an inconvenience
    in their application)

    So, some variant of C will be peeled off and maintained by a separate
    standards organization. This will force compiler writers to have
    to support a second language (this is already in the works with
    MISRA, C*, etc.). How humiliating for a group charged with promoting
    a language to see their market head off in a different direction...
    BECAUSE of their actions!

    The same bit of source code written in a particular language
    should result in the same general behavior regardless of processor
    or compiler choice.

    It will, if it is defined code.

    No. Even defined code gives the compiler vendor extra license. The
    code for:

    x = 5;
    print("hello");

    need not perform those statements in that order. So, I can compile
    it on compiler A and come up with a different "sequence of operations"
    (I'm being generous and giving them freedom in the details) than
    on compiler B.

    If I am *debugging* that bit of code, I most definitely care which
    order the (ASM-associated) statements are executed, even if there is not
    an implicit dependency *in* the source.

    x = 5;
    ...
    x += x;

    can be replaced. And, if x is never referenced, all of that code can
    be elided. And the compiler is under no obligation to TELL me that it
    has done these things! ("Oh! I meant *y*, not *x*!")

    Remember who the customer is.

    That contract was broken by C more than 20 years ago.

    C is now an academic exercise and playground for c.l.c types.

    All of my colleagues use C89 or K&R -- because we have to be able
    to PROVE that the source code that the customer receives and
    systematically reviews corresponds with what is happening in the
    hardware.

    Use of K&R C for new code looks irresponsible to me: checking that
    code accepted by K&R compiler has no bugs is more effort than

    You aren't limited to using a K&R compiler to check the code's
    syntax. E.g., I make extensive use of KLEE which builds on LLVM
    /to find bugs/. But, use K&R/C89 to generate PRODUCTION code
    that can be more readily validated.

    the same for modern C. C89 looks reasonable, but it gains you
    a little. Main thing are compilers. There are two mainstream
    ones with multiple targets, that is gcc and clang. There were
    few respectable compilers from specific vendors targeting their
    targets, but most seem to be gone now. And all compilers beside
    gcc and clang seem to get much less use and consequently are
    less tested.

    More use doesn't translate into better quality.

    Remember, the goal is for the PROGRAMMER to be able to get his
    algorithm to run on targeted hardware. The compiler loses
    all significance once the code is running. And, can be
    replaced by a competing product (or development methodology)
    if it proves to be a hindrance to that goal.

    When using C compiler you may have trouble for various reasons.
    One thing are ordinary bugs in C code. Other thing are bugs
    which cause undefined behaviour. Third thing are compiler bugs,
    when source code is well defined, but compiler produces wrong
    object code. AFAICS modern compilers are much more helpful
    with catching ordinary bugs. Concerning wrong code, they

    But, they are under no obligation to do so! As such, all
    of that "help" can go away in the next release.

    They also go too far in trying to treat standard library
    functions as "operators" to be optimized like any other
    statements. E.g., I use NONE of the standard library function
    names because I don't want the compiler second-guessing my intent.
    Trust the programmer to be competent. He doesn't need a "nanny".

    seem to be better, but larger size of compiler and increased
    complexity may leave some doubts. Undefined behaviour could
    catch you even 35 years ago: at that time users complained
    that their working code is broken by new C compilers. Users
    learned that old code was wrong and how to write correct
    code. I do not see much difference now: you learn to write
    code in a different way, which has additional advantage that
    code avoinding risky constructs usually is also clearer.

    But you aren't *instructed* -- the compiler is under no obligation
    to point out what you are doing wrong. Instead, it can just
    elide (or alter) your efforts with license.

    I listened to a Mexican gentleman speaking about his wife, the
    other day. He kept using the pronoun "he/him" in every reference.
    I could decide that I could ignore everything he was saying because
    his grammar was incorrect -- that would surely be the easiest way
    to interpret his comments. Or, I can expend a bit of effort and
    make the mental translation of he/him to she/her.

    Apparently, compiler writers are too lazy to even make such an attempt.

    [I could not correct him as I had no means of contacting him]

    Any "optimizations" have to be relatively trivial and intuitive
    and visible at the preprocessor stage; the preprocessed output
    should accuraely define what the code ACTUALLY does, not what the
    developer HOPES it will do (subject to the compiler-writer's
    whims).

    You preach specifications yet you have trouble that compiler also
    have specification. Stay withing specified (defined) constructs
    and your code will have expected effect. If you expect more
    than is specified, then you expect too much.

    Because the specifications adopted for the compilers are nonsense
    from the standpoint of the compiler's user!

    C'mon, I'm not the only one singing this song:

    <https://www.yodaiken.com/2021/05/19/undefined-behavior-in-c-is-a-reading-error/>
    <https://www.yodaiken.com/2021/05/16/c-is-not-a-serious-programming-language/>
    <https://www.kb.cert.org/vuls/id/162289/>
    <http://www.complang.tuwien.ac.at/papers/ertl17kps.pdf>
    <https://groups.google.com/g/boring-crypto/c/48qa1kWignU/m/o8GGp2K1DAAJ>
    <http://homes.cs.washington.edu/~akcheung/getFile.php?file=apsys12.pdf>

    The standards committee has placed an emphasis on "compiler tricks"
    at the expense of maintaining a usable language.

    And, falling back on standards language to justify their excesses.

    Concerning "what the code ACTUALLY does", I expect compiler
    to delete checks that are known to pass at compile time
    (such checks may be generated from macros and I want to keep
    my macros). I expect compiler to expand inline simple functions,
    when it leads to smaller code size. In general generated
    code may look quite different than expectation of some
    folks. What counts is the effect.

    What counts is whether or not the code faithfully implements the
    algorithm created by the developer. Because the developer will be TESTING
    and DEBUGGING to those expectations, not the compiler writer's view of
    what he *should* have written.

    For delay loop I have assembly code: C compiler may and should
    collapse to nothing loops that take time but otherwise do not
    contribute to the result. And I find atempts to cripple
    optimization in order to preserve delay loop to be too
    unreliable. But otherwise C allows me to write what
    should be done and compiler generates appropriate code.

    Loop unrolling, function inlining, constant folding, constant
    subexpression elimination, etc. -- all things that are OBVIOUS to
    someone inspecting the source ("Yes, I understand that STRINGLENGTH+1
    is intended to account for the NUL termination in a more visible manner"
    or "OK, I see 3+3+4+1 is a good way of representing a telephone number"
    or "Yeah, I guess you could do those steps N (constant) times in a row
    instead of wrapping them in a loop that is executed N times")

    In that way, the "source" is a writing shorthand but one that still
    completely constrains the implementation.

    E.g., I use a construct:
    object=>method(arguments)
    liberally in my code. The language doesn't support it (of course).
    But, it is known to be a representational shorthand -- something to
    unclutter the code and highlight the actual operation being performed.

    There is no surprise when you look at the preprocessor output...
    for ANY instance of such an expression.

    Even the optimizing IDL compiler is obvious in the "optimizations"
    that it makes (e.g., choosing effective marshalling sequences instead
    of leaving that up to the developer)

    That is slightly different thing. Optimizations require some effort.
    If you have preprocessor/precompiler it makes no sense to duplicate optimizations done by C compiler. Only things that C compiler
    can not optimize are worth optimization in a precompiler and only
    when they are important enough. More optimizations in C compiler
    means that your precompiler may be simpler. For example
    precompiler may naively insert code trusting that C compiler
    will remove it in cases where it is not needed.
    The compiler understands the target hardware. The preprocessor
    (assuming you are trying to write portable code) only understands
    the language syntax and the application.

    So, you promote the optimizations that the developer would freely
    embrace to the sources via the preprocessor as a bridge between what
    the developer wants to say and what the code should realistically do
    (e.g., char SSN[3 +1+ 2 +1+ 4 +1] is much cleaner to write than
    SSN[12] or, gasp, SSN[11])

    Expecting the developer to write code that understands and exploits
    how the processor issues opcodes defeats the purpose of having a HLL.
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Fri Nov 13 04:58:44 2026
    From Newsgroup: sci.electronics.design

    On 11/12/2026 4:55 AM, C||il|!n Niocl|is|!n Glost|-ir wrote:
    Don Y <blockedofcourse@foo.invalid> wrote: |--------------------------------------------------------------|
    |"Perhaps we should let hardware people write code and publicly|
    |critique their failures? Or, let software people design the |
    |hardware and let the hardware folks write the code??" | |--------------------------------------------------------------|

    VHDL
    (S. HTTP://Gloucester.Insomnia247.NL/ fuer Kontaktdaten!)

    It does level the playing field, but if you write VHDL as if you're
    doing software, you kind of fail. At least, that's what I got from the
    free VHDL book I found online a few years ago.

    Personally, I think it's very instructive to design and implement a
    circuit using 74xx logic and a breadboard. At least, that hammers home
    that you're designing hardware, and not just writing software.

    There are some more intricacies involved, such as -- and I never exper-
    ienced this -- is wires so close together they become capacitors, or
    something along those lines. It was just one of these off hand remarks
    by the teachers in my /coding school/. And then I never took any of the
    more advanced electronics classes, as they weren't in my curriculum.

    So I have a lot to catch up on, if I'm to do this for real.
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Fri Nov 13 05:04:10 2026
    From Newsgroup: sci.electronics.design

    On 11/12/2026 12:35 PM, Bill Sloman wrote:
    On 10/11/2026 10:02 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 5:14 AM, Waldek Hebisch wrote:
    Actually, at '-O2' version of 'gcc' that I use replaces explicit
    division by 0 by the 'ud2' instruction.

    Shouldn't that be considered /erroneous compilation/?-a I mean, you can
    write a signal handler that catches all signals, and isn't /division by
    zero/ one of the Posix signals?-a Is the compiler prescient and makes
    sure this only happens if the binary is ever run when the user isn't ex-
    plictly expecting a division by zero -- rather than /invalid opcode/?

    Does the question bother you?-a Am I too high, or too low on the /moron/
    vs. /jedi master/ axis to be understood?-a Should we put all the GCC devs
    in therapy for believing they are prescient?

    I imagine that there are situations where the programmer can be
    confident that the divisor is never going to be zero.

    A cautious programmer might test for a zero divisor and run an error recovery routine rather than doing the division.

    To me it does look as if you've posted an under-informed response.


    You should try to write a signal handler that looks at the offending instruction or something. It's been a while since I did this, and that
    was for -- overflow? -- I was experimenting with my own + operator in
    C++ and injected the INTO instruction -- that was before AMD in their never-ending wisdom removed it from the 64bit instruction set -- which I
    then handled in a fancy Posix signal handler, and if I recall correctly
    I had access to the very instruction that caused the "fault" and could
    use it to redirect to a C++ exception mechanism as if that happened in
    the function causing the "fault" or "trap" or whatever Intel wants to
    call it.

    Anyway, that code is long lost, and I never did much with it. It was an illuminating exercise though, and taught me that there's a lot a com-
    piler can do.


    Best wishes, and happy Posix signal handlers!
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Fri Nov 13 05:16:07 2026
    From Newsgroup: sci.electronics.design

    On 11/12/2026 8:58 PM, Don Y wrote:
    On 11/11/2026 2:50 PM, Don Y wrote:
    You have to THINK about the problem instead of just naively
    assuming everything is important.

    When deadlines are soft, the "value" of a task's completion
    is no longer bi-valued (value vs. no value at all) but, rather,
    some potentially complex function of lateness.-a Your goal,
    then, is to sort out how to shuffle resources to "maximize value"
    (across the set of tasks) instead of the all-or-nothing decision
    that HRT suggests.

    An SRT design tends to embed more intelligence in the
    goal setting and resource management.-a You *shed*
    tasks for which resource expenditure provides little
    *value* at the possible expense of other MORE valuable
    tasks.
    Here's a simple, practical example that illustrates the costs and
    value of a single "task" and how it might be instantiated:

    E.g., if I want to "water the citrus", I need the hardware interface to the irrigation valves that is present on the irrigation controller.-a Otherwise, I can't "gate" water to that portion of the yard.-a I need the "grey water" supply and its metering hardware (another node) to source the water.-a And, possibly, the municipal water supply and its metering hardware -- in the event that-a there is insufficient grey water available.

    I need the database that tracks weather history so I can determine how *much*
    water will be required (based on precipitation, temperatures, wind, etc.).
    If it rained recently -- or, if it has been hot/windy/dry -- then the current
    needs of those plants will differ.

    And, I need memory and "MIPS-seconds" on each of these nodes to actually do the "work" required.-a So, those devices must be powered up and have sufficient
    "spare" resources to host the activities that I will require of them.

    During daylight hours, I'll likely have a surplus of electrical power (assuming the power system's estimate of the remaining available solar
    power will suffice to top of the storage batteries).-a At night, power
    may be in shorter supply (or, "more costly" so this has to be justified
    by the "value" of getting the task accomplished!).-a OTOH, irrigation is
    more effective at night as evaporative losses are greatly reduced and
    the flora will be most receptive to water uptake as the sun rises.
    OToOH, if the power has failed, then power will be even more precious; perhaps risk the fruit taking a hit so some more important task can be accomplished?

    The scheduler now has to find a time when those resources are available
    (or, can be MADE available -- powering hardware on if presently off,
    evicting jobs of lower priority, etc.) and then admit the task to the
    run queue.

    *Hoping* that conditions don't change significantly before it finishes
    (a minute from now, an occupant may want to take a shower and the drop
    in water pressure caused by the huge load represented by the citrus
    might be objectionable -- so, that task will have to be paused or
    aborted)

    If I had taken the attitude that "everything is important and of immediate need", then I would have had to increase the size of the grey water tank
    (and the pump and piping that distributes its contents as I might have had
    to irrigate EVERYTHING on the property at the same instant!), the diameter
    of the municipal water main to handle those cases where the grey water
    supply was inadequate, the size of the storage battery in case lots of things
    wanted to happen "after dark", the number and capabilities of the CPUs
    doing the work, etc.

    [Irrigation is much easier for people to wrap their heads around than
    the effort and resources required to train a speech model or update a
    scene analysis, hence my reliance on it as an example]

    This, of course, is just stupid.-a A smarter approach is to juggle priorities
    so you can do everything with *less*.-a By recognizing that not everything has to happen immediately.-a By treating most "deadlines" as ADVISORY and
    not HARD REQUIREMENTS.

    [You can see there are far more design decisions involved in this
    approach!]

    Thank you. My SO has plants as a hobby, but I have not thought to
    create automated irrigation systems, so this example hits perfectly
    home.

    One thing you can add is the failure mode: when the irrigation system completely gives up, it can decide to /phone home/ and irritate a human.

    That said, it's also best to keep a mark-one-eyeball on the growing
    plants, and not be completely dependent upon an automated system that
    can decide to fail silently. Such as if someone steals the battery.


    Best wishes, and happy irrigation systems!
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Don Y@blockedofcourse@foo.invalid to sci.electronics.design on Thu Nov 12 14:50:54 2026
    From Newsgroup: sci.electronics.design

    On 11/12/2026 2:16 PM, Johann 'Myrkraverk' Oskarsson wrote:
    [Irrigation is much easier for people to wrap their heads around than
    the effort and resources required to train a speech model or update a
    scene analysis, hence my reliance on it as an example]

    This, of course, is just stupid.-a A smarter approach is to juggle priorities
    so you can do everything with *less*.-a By recognizing that not everything >> has to happen immediately.-a By treating most "deadlines" as ADVISORY and
    not HARD REQUIREMENTS.

    [You can see there are far more design decisions involved in this approach!]

    Thank you.-a My SO has plants as a hobby, but I have not thought to
    create automated irrigation systems, so this example hits perfectly
    home.

    Irrigation (for folks who don't want "natural" landscapes) is essential
    here as we don't get much precipitation. If you don't water it (whatever
    "it" happens to be), then it likely dies.

    Except the weeds, of course!

    One thing you can add is the failure mode: when the irrigation system completely gives up, it can decide to /phone home/ and irritate a human.

    My goal is to address the needs of people later in life who may not
    *remember* to irrigate -- which would lead to plants needing to be
    replaced or removed and a decrease in the value of the property.

    Not to mention the value to the resident.

    That said, it's also best to keep a mark-one-eyeball on the growing
    plants, and not be completely dependent upon an automated system that
    can decide to fail silently.-a Such as if someone steals the battery.

    No batteries. But, there are many failure modes that you can
    detect (if so instrumented).

    E.g., I watch the rate of water consumption while irrigating.
    No idea what it *should* be (numerically) BUT it should be
    very similar to what it was the last time this "zone" was
    irrigated. An increase could represent an emitter that
    has come off (allowing water to flow at a higher rate than
    prescribed) or a failed distribution line. A decrease
    could mean a clogged emitter or blocked distribution line.

    A traditional (non-instrumented) irrigation system would
    rely on the user (resident) to "notice" these things, either
    by noting unexpected pools of water (or "gushers") or
    plants that are dried out (or overwatered). Expecting a
    resident to be capable of noticing these things *and* to
    actually notice them time can be difficult for folks later
    in life.

    In either case, "service" is required (without relying on the
    resident to perform that action). So, call for it directly or
    advise a "caretaker" to do so.

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Fri Nov 13 06:19:02 2026
    From Newsgroup: sci.electronics.design

    On 11/12/2026 10:31 PM, Bill Sloman wrote:
    In any case, trying to use /knowledge/ as a status symbol seems quite
    stupid to me; yet a lot of people behave this way.

    Correcting errors isn't "using knowledge as a status symbol", except in
    the sense that being able to correct errors does involve asserting a superior status, but since the alternative is to leave errors unremarked
    and uncorrected, it is also taking responsibility.

    Of course you think so, but leaving errors is fine too. After all, I'm
    not being paid to educate you on the errors of your ways.
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Don Y@blockedofcourse@foo.invalid to sci.electronics.design on Thu Nov 12 16:14:47 2026
    From Newsgroup: sci.electronics.design

    On 11/12/2026 1:58 PM, Johann 'Myrkraverk' Oskarsson wrote:
    It does level the playing field, but if you write VHDL as if you're
    doing software, you kind of fail.-a At least, that's what I got from the
    free VHDL book I found online a few years ago.

    Hardware works in parallel. Most software people think serially;
    this is why multithreading and multiprocessing are such stumbling
    blocks for neophytes.

    But, hardware makes things like hazzards and races more obvious.
    When you encounter them in your software, you can understand
    their analogs if you have a hardware background.

    Likewise, atomic operations.

    Personally, I think it's very instructive to design and implement a
    circuit using 74xx logic and a breadboard.-a At least, that hammers home
    that you're designing hardware, and not just writing software.

    The problem with hardware is that it IS hardware. You actually need
    physical devices if you want to BUILD a prototype.

    The advantage to VHDL and other simulation technologies is you can
    *emulate* the hardware to see how it works (as well as how it fails
    to work) without having to invest all that time and hassle.

    There's nothing worse than being handed a prototype that someone
    built incorrectly and you now have to debug their assembly errors
    before you can examine the behavior of your circuit.

    There are some more intricacies involved, such as -- and I never exper- ienced this -- is wires so close together they become capacitors, or something along those lines.-a It was just one of these off hand remarks
    by the teachers in my /coding school/.-a And then I never took any of the more advanced electronics classes, as they weren't in my curriculum.

    When I was working with large wirewrap panels (think 10+ square feet
    per panel), we also had to worry about "cold flow" if the wires were
    pulled too tight as the panel was wrapped. For "hand wrapped" panels,
    you had to rely on the skillset of the technician.

    [I still have my two WW guns, slit-n-wrap bits, no-niks, needle-nose
    strippers (worth their weight in gold!) and embroidery hooks!]

    So I have a lot to catch up on, if I'm to do this for real.

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Fri Nov 13 16:35:10 2026
    From Newsgroup: sci.electronics.design

    On 13/11/2026 9:19 am, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/12/2026 10:31 PM, Bill Sloman wrote:
    In any case, trying to use /knowledge/ as a status symbol seems quite
    stupid to me; yet a lot of people behave this way.

    Correcting errors isn't "using knowledge as a status symbol", except
    in the sense that being able to correct errors does involve asserting
    a superior status, but since the alternative is to leave errors
    unremarked and uncorrected, it is also taking responsibility.

    Of course you think so, but leaving errors is fine too.

    You may think so but you aren't any kind of exemplary character.

    -aAfter all, I'm not being paid to educate you on the errors of your ways.

    What you post isn't exactly educational. It's sort of useful as an
    example of how silly you can get if you don't know what you are talking
    about, but we aren't short of those kinds of posts here.
    --
    Bill Sloman, Sydney
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Fri Nov 13 16:39:37 2026
    From Newsgroup: sci.electronics.design

    On 13/11/2026 8:04 am, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/12/2026 12:35 PM, Bill Sloman wrote:
    On 10/11/2026 10:02 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 5:14 AM, Waldek Hebisch wrote:
    Actually, at '-O2' version of 'gcc' that I use replaces explicit
    division by 0 by the 'ud2' instruction.

    Shouldn't that be considered /erroneous compilation/?-a I mean, you can
    write a signal handler that catches all signals, and isn't /division by
    zero/ one of the Posix signals?-a Is the compiler prescient and makes
    sure this only happens if the binary is ever run when the user isn't ex- >>> plictly expecting a division by zero -- rather than /invalid opcode/?

    Does the question bother you?-a Am I too high, or too low on the /moron/ >>> vs. /jedi master/ axis to be understood?-a Should we put all the GCC devs >>> in therapy for believing they are prescient?

    I imagine that there are situations where the programmer can be
    confident that the divisor is never going to be zero.

    A cautious programmer might test for a zero divisor and run an error
    recovery routine rather than doing the division.

    To me it does look as if you've posted an under-informed response.


    You should try to write a signal handler that looks at the offending instruction or something.

    I'd have to lot of work before I'd be in a position to do that, and
    nobody would hire me to do that kind of work. I'm a much better hardware engineer than I am a programmer.

    <snip>
    --
    Bill Sloman, Sydney
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Fri Nov 13 14:47:29 2026
    From Newsgroup: sci.electronics.design

    On 11/12/2026 10:39 AM, Bill Sloman wrote:
    On 12/11/2026 11:50 am, john larkin wrote:
    On Wed, 11 Nov 2026 14:54:11 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 11/11/2026 2:42 am, john larkin wrote:
    On Tue, 10 Nov 2026 18:38:47 +0800, Johann 'Myrkraverk' Oskarsson
    <johann@myrkraverk.invalid> wrote:

    On 11/10/2026 6:12 PM, Liz Tuddenham wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:


    The saving grace of IQ testing is that if you good in one area you tend
    to be good in others - not perhaps as exceptionally good, but still
    better than average. It may be that what is being tested for is more the >>> absence of actual brain damage rather than some kind of superior
    processing.

    The specialized-genius bit can also make them obnoxious.

    Not in my experience.

    Sometimes you actually, probably accidentally, approach humor.

    Back in about 1968 I wrote a few scripts for the Melbourne University Architects Revue. Some of them got into the revue and got laughs. Rowan Atkinson did much better when he was at Cambridge, and dropped his
    master's degree in electronics in favour of a theatrical career.

    Your own credentials as a judge of humour are even less impressive.

    There were a few specialised genius types around EMI Central Research
    when I worked there, and only one of them was obnoxious. It wasn't his genius that was the problem, but his irresponsible antics.

    William S. Percival was probably cleverer, but quite easy to get on with.

    https://en.wikipedia.org/wiki/Distributed_amplifier


    Humor is just one of these things that's easy because it's natural for
    humans to do. I wrote a little bit of it and published on Usenet ear-
    lier this year. Look for /Dad jokes of alt.folklore.computers/ if you
    care to read it.

    And writing a good joke is a tangential skill of telling one. You start
    and write it, and the first attempt isn't really good, and then you
    start again with an empty page, and write the whole thing from the top
    in a slightly different way, and if it's good enough you can show it off
    or you junk it, and do it for the third time.

    Here you can see my second attempt,

    https://bsky.app/profile/myrkraverk.bsky.social/post/3mz6fwrupz22e

    and it's "good enough" to share with the world, and I haven't written
    this one for the third time. Do you think I should? Or is it /good
    enough/ for a smirk?

    It's a three minute video, where I'm typing it in, so you can either
    watch or skip to the last second and pause.

    The point of all of this? Comedy is fun, you should try it. It's mere-
    ly a matter of practice.


    Best wishes, and happy humor!
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Fri Nov 13 17:54:57 2026
    From Newsgroup: sci.electronics.design

    On 13/11/2026 7:58 am, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/12/2026 4:55 AM, C||il|!n Niocl|is|!n Glost|-ir wrote:
    Don Y <blockedofcourse@foo.invalid> wrote:
    |--------------------------------------------------------------|
    |"Perhaps we should let hardware people write code and publicly|
    |critique their failures?-a Or, let software people design the-a |
    |hardware and let the hardware folks write the code??"-a-a-a-a-a-a-a-a |
    |--------------------------------------------------------------|

    VHDL
    (S. HTTP://Gloucester.Insomnia247.NL/ fuer Kontaktdaten!)

    It does level the playing field, but if you write VHDL as if you're
    doing software, you kind of fail.-a At least, that's what I got from the
    free VHDL book I found online a few years ago.

    Personally, I think it's very instructive to design and implement a
    circuit using 74xx logic and a breadboard.-a At least, that hammers home
    that you're designing hardware, and not just writing software.

    74xx logic sucks. Do it in ECL or ECLinPS and you might end up with
    something fast enough to be interesting.

    There are some more intricacies involved, such as -- and I never
    experienced this -- is wires so close together they become capacitors,
    or something along those lines.

    This shows up in analog circuit layout, and I've been finding and fixing
    those kind of layout faults since the 1970's.

    Coupling can be inductive as well as capacitative and long parallel
    tracks can act as transmission line transformers.

    It was just one of these off hand remarks
    by the teachers in my /coding school/.-a And then I never took any of the more advanced electronics classes, as they weren't in my curriculum.

    So I have a lot to catch up on, if I'm to do this for real.

    Quite a lot.
    --
    Bill Sloman, Sydney

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Fri Nov 13 18:01:54 2026
    From Newsgroup: sci.electronics.design

    On 13/11/2026 10:14 am, Don Y wrote:
    On 11/12/2026 1:58 PM, Johann 'Myrkraverk' Oskarsson wrote:
    It does level the playing field, but if you write VHDL as if you're
    doing software, you kind of fail.-a At least, that's what I got from the
    free VHDL book I found online a few years ago.

    Hardware works in parallel.-a Most software people think serially;
    this is why multithreading and multiprocessing are such stumbling
    blocks for neophytes.

    But, hardware makes things like hazzards and races more obvious.
    When you encounter them in your software, you can understand
    their analogs if you have a hardware background.

    Likewise, atomic operations.

    Personally, I think it's very instructive to design and implement a
    circuit using 74xx logic and a breadboard.-a At least, that hammers home
    that you're designing hardware, and not just writing software.

    The problem with hardware is that it IS hardware.-a You actually need physical devices if you want to BUILD a prototype.

    LTSpice lets you simulate a lot pf potential problems.

    The advantage to VHDL and other simulation technologies is you can
    *emulate* the hardware to see how it works (as well as how it fails
    to work) without having to invest all that time and hassle.

    There's nothing worse than being handed a prototype that someone
    built incorrectly and you now have to debug their assembly errors
    before you can examine the behavior of your circuit.

    There are some more intricacies involved, such as -- and I never
    experienced this -- is wires so close together they become capacitors, or
    something along those lines.-a It was just one of these off hand remarks
    by the teachers in my /coding school/.-a And then I never took any of the
    more advanced electronics classes, as they weren't in my curriculum.

    When I was working with large wirewrap panels (think 10+ square feet
    per panel), we also had to worry about "cold flow" if the wires were
    pulled too tight as the panel was wrapped.-a For "hand wrapped" panels,
    you had to rely on the skillset of the technician.

    [I still have my two WW guns, slit-n-wrap bits, no-niks, needle-nose strippers (worth their weight in gold!) and embroidery hooks!]

    That's a bit dark ages now. You can get to printed circuit boards very
    rapidly now, if you have the money and in any commercial project you will.

    So I have a lot to catch up on, if I'm to do this for real.

    Lots.
    --
    Bill Sloman, Sydney


    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Fri Nov 13 15:22:18 2026
    From Newsgroup: sci.electronics.design

    On 11/13/2026 5:50 AM, Don Y wrote:
    On 11/12/2026 2:16 PM, Johann 'Myrkraverk' Oskarsson wrote:
    [Irrigation is much easier for people to wrap their heads around than
    the effort and resources required to train a speech model or update a
    scene analysis, hence my reliance on it as an example]

    This, of course, is just stupid.-a A smarter approach is to juggle
    priorities
    so you can do everything with *less*.-a By recognizing that not
    everything
    has to happen immediately.-a By treating most "deadlines" as ADVISORY and >>> not HARD REQUIREMENTS.

    [You can see there are far more design decisions involved in this
    approach!]

    Thank you.-a My SO has plants as a hobby, but I have not thought to
    create automated irrigation systems, so this example hits perfectly
    home.

    Irrigation (for folks who don't want "natural" landscapes) is essential
    here as we don't get much precipitation.-a If you don't water it (whatever "it" happens to be), then it likely dies.

    Except the weeds, of course!

    One thing you can add is the failure mode: when the irrigation system
    completely gives up, it can decide to /phone home/ and irritate a human.

    My goal is to address the needs of people later in life who may not *remember* to irrigate -- which would lead to plants needing to be
    replaced or removed and a decrease in the value of the property.

    Not to mention the value to the resident.

    That said, it's also best to keep a mark-one-eyeball on the growing
    plants, and not be completely dependent upon an automated system that
    can decide to fail silently.-a Such as if someone steals the battery.

    No batteries.-a But, there are many failure modes that you can
    detect (if so instrumented).

    E.g., I watch the rate of water consumption while irrigating.
    No idea what it *should* be (numerically) BUT it should be
    very similar to what it was the last time this "zone" was
    irrigated.-a-a An increase could represent an emitter that
    has come off (allowing water to flow at a higher rate than
    prescribed) or a failed distribution line.-a A decrease
    could mean a clogged emitter or blocked distribution line.

    A traditional (non-instrumented) irrigation system would
    rely on the user (resident) to "notice" these things, either
    by noting unexpected pools of water (or "gushers") or
    plants that are dried out (or overwatered).-a Expecting a
    resident to be capable of noticing these things *and* to
    actually notice them time can be difficult for folks later
    in life.

    In either case, "service" is required (without relying on the
    resident to perform that action).-a So, call for it directly or
    advise a "caretaker" to do so.

    Dear Don,

    Thank you for that exposition. You said there were no batteries, and I
    want to add real life experience to your /threat model/. Here in this
    part of the world, it's common to cut off the electricity if the condo-
    minium dues are not paid on time, and if that happens during an extended vacation -- which did happen to me -- it can be several weeks, or months
    until someone notices the power is off.

    And during that time, the plants would die, with the irrigation system
    unable to even phone home to irritate a human.

    For your given target market, that can also happen if a /helpful family
    member/ comes along and turns off the breaker to save a few dollars dur-
    ing an extended vacation away from home.

    Now, whether people are willing to pay for the extra expense of a bat-
    tery addition is another matter, but you can always add such stories to
    your sales spiel and see what happens.


    Best wishes, and happy irrigation systems!
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Don Y@blockedofcourse@foo.invalid to sci.electronics.design on Fri Nov 13 01:33:51 2026
    From Newsgroup: sci.electronics.design

    On 11/13/2026 12:22 AM, Johann 'Myrkraverk' Oskarsson wrote:
    Thank you for that exposition.-a You said there were no batteries, and I
    want to add real life experience to your /threat model/.-a Here in this
    part of the world, it's common to cut off the electricity if the condo- minium dues are not paid on time, and if that happens during an extended vacation -- which did happen to me -- it can be several weeks, or months until someone notices the power is off.

    And during that time, the plants would die, with the irrigation system
    unable to even phone home to irritate a human.

    All of my "nodes" are fed from a central power supply via PoE.
    So, if it is unavailable, NOTHING is unavailable (HVAC controls,
    surveillance, security system, etc.). I.e., one expects power
    to be available or the occupant has bigger problems to address
    regarding "livability" in the residence.

    For your given target market, that can also happen if a /helpful family member/ comes along and turns off the breaker to save a few dollars dur-
    ing an extended vacation away from home.

    It is highly unlikely that such a thing would happen. There is
    a fixed cost associated with having a power line connected to your
    home, regardless of the amount of energy consumed. If the house was
    empty, the loads powered would be those that the house/occupant
    considered essential. Recall, many homes have indoor plants, pets, etc.
    so it is unlikely that a house will be "shuttered" for long periods.

    Even our "winter visitors" (absent for 6+ months out of every year)
    find it easier to leave the electric service connected than to worry about
    any savings that might acrue from turning it off.

    Now, whether people are willing to pay for the extra expense of a bat-
    tery addition is another matter, but you can always add such stories to
    your sales spiel and see what happens.
    The problem with batteries is that they require maintenance. And,
    have sorely limited capacities. E.g., running a security camera
    (PTZ with Ir illuminators) off a battery would quickly deplete it.
    This also would limit what could be done (computationally) in each
    such node as the batteries would also have to power any "smarts"
    colocated therein.

    So, *which* cameras should be battery powered? Which appliances
    (irrigation system, HVAC, security, etc.)?

    [I have ~300 nodes that could potentially be powered. Or not.]

    The simpler solution is to let the user configure the system
    for which loads it wants to keep powered -- from the PoE PSE.
    E.g., it is unlikely he would want TV broadcasts recorded in his
    extended absence so no need to power the TV tuners or the media tank.
    Should the telephone interface remain active? Or, should it "go silent"?
    (What if the resident wants to telephone the system to effect some change?) Should the surveillance cameras monitor for wildlife in the yard?
    Intruders around the residence? etc.

    The preferences of one residence may differ from another.
    And, may change over time.

    So, put the "battery" at the PSE and let it decide what to
    support in the absence of mains power.

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Martin Brown@'''newspam'''@nonad.co.uk to sci.electronics.design on Fri Nov 13 09:21:52 2026
    From Newsgroup: sci.electronics.design

    On 08/11/2026 19:27, Waldek Hebisch wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:
    On 7/11/2026 3:57 pm, john larkin wrote:
    On Sat, 7 Nov 2026 15:38:13 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 7/11/2026 5:54 am, john larkin wrote:

    Being a compiler is tedious.

    Unlike some architectures that I might name.

    A language isn't an architecture. It has to be processed into something >>>> that can be run on hardware (which will have an architecture - it is
    constructed from components that have been connected by some kind of
    architect).

    If that post were compiled to assembly language, it would be a nop.

    As a compiler, you do seem to be sub-optimal, if not downright
    afunctional. Ignoring implications you don't like isn't an option for a
    compiler. There can be functions they can't parse - in your case don't
    want to parse - but that doesn't generate a nop.

    You clearly go into area were you are not an expert. Google a
    bit about "undefined behaviour" in C and you will learn that
    when source code makes no sense, then C compier can do
    essentially anything, which in not so rare cases means that
    compiler simply deletes what programmer wrote.

    It can also do that at high levels of optimisation if the code is
    correct as intended but has no side effects beyond its own scope.

    The most common UB error in C is stupid buffer overrun followed by
    signed vs unsigned integer arithmetic overflow which can catch almost
    anyone out on a bad day. It only takes a moments lapse in concentration.

    Explicit nop-s are rarely emited by compilers,

    Now you are showing *your* ignorance. Modern optimising compilers often
    issue variable length noops to pad hot branch target destinations where
    they need to be correctly aligned for optimum use of instruction cache.

    but one can get 'return'
    (if compiler decides that garbage source is doing nothing
    useful, but it will eventually finish) or executing whatever
    happens to be after deleted code (if compiler decides that
    code should be never executed because it performed illegal
    operation).

    I always liked the HCF instruction (Halt and Catch Fire) for undefined behaviour (there is *nothing* in the standard to say it can't do that).
    --
    Martin Brown

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Fri Nov 13 22:05:46 2026
    From Newsgroup: sci.electronics.design

    On 13/11/2026 5:47 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/12/2026 10:39 AM, Bill Sloman wrote:
    On 12/11/2026 11:50 am, john larkin wrote:
    On Wed, 11 Nov 2026 14:54:11 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 11/11/2026 2:42 am, john larkin wrote:
    On Tue, 10 Nov 2026 18:38:47 +0800, Johann 'Myrkraverk' Oskarsson
    <johann@myrkraverk.invalid> wrote:

    On 11/10/2026 6:12 PM, Liz Tuddenham wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:

    <snip>

    The point of all of this?-a Comedy is fun, you should try it.-a It's
    merely a matter of practice.

    Comedy can be fun, and I've done a bit of it. Claiming that it is just a matter of practice means that you don't know what you are talking about,
    as usual.

    <snip>
    --
    Bill Sloman, Sydney
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From =?UTF-8?B?Q8OzaWzDrW4=?= =?UTF-8?B?IE5pb2Nsw6Fzw61u?= =?UTF-8?B?IEdsb3N0w6lpcg==?=@thanks-to@Taf.com to sci.electronics.design on Fri Nov 13 12:35:36 2026
    From Newsgroup: sci.electronics.design

    All my stolen phones became stolen indoors - one of them became stolen
    in a home (when I was away).
    (S. HTTP://Gloucester.Insomnia247.NL/ fuer Kontaktdaten!)
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Fri Nov 13 20:49:18 2026
    From Newsgroup: sci.electronics.design

    On 11/13/2026 7:05 PM, Bill Sloman wrote:
    On 13/11/2026 5:47 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/12/2026 10:39 AM, Bill Sloman wrote:
    On 12/11/2026 11:50 am, john larkin wrote:
    On Wed, 11 Nov 2026 14:54:11 +1100, Bill Sloman <bill.sloman@ieee.org> >>>> wrote:

    On 11/11/2026 2:42 am, john larkin wrote:
    On Tue, 10 Nov 2026 18:38:47 +0800, Johann 'Myrkraverk' Oskarsson
    <johann@myrkraverk.invalid> wrote:

    On 11/10/2026 6:12 PM, Liz Tuddenham wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:

    <snip>

    The point of all of this?-a Comedy is fun, you should try it.-a It's
    merely a matter of practice.

    Comedy can be fun, and I've done a bit of it. Claiming that it is just a matter of practice means that you don't know what you are talking about,
    as usual.

    <snip>


    Dear Bill,

    Now you're so wrong you have egg on your face. How does it feel? Try-
    ing to claim comedy isn't about practice is just your coping mechanism
    for being so unfunny in life everyone stopped talking to you.

    So don't project your coping mechanism on other people, it's not
    healthy.


    Best wishes, and happy comedies!
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Fri Nov 13 20:51:25 2026
    From Newsgroup: sci.electronics.design

    On 11/13/2026 8:35 PM, C||il|!n Niocl|is|!n Glost|-ir wrote:
    All my stolen phones became stolen indoors - one of them became stolen
    in a home (when I was away).
    (S. HTTP://Gloucester.Insomnia247.NL/ fuer Kontaktdaten!)

    Well, that's a bummer. It suggests you don't have a stolen phone prob-
    lem, but a /home problem/. I suggest you move out immediately, and find
    a new place to live. For some value of immediately, I don't know your financial situation.


    Best wishes, and good luck finding a new home!
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Sat Nov 14 01:16:43 2026
    From Newsgroup: sci.electronics.design

    On 13/11/2026 11:49 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/13/2026 7:05 PM, Bill Sloman wrote:
    On 13/11/2026 5:47 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/12/2026 10:39 AM, Bill Sloman wrote:
    On 12/11/2026 11:50 am, john larkin wrote:
    On Wed, 11 Nov 2026 14:54:11 +1100, Bill Sloman <bill.sloman@ieee.org> >>>>> wrote:

    On 11/11/2026 2:42 am, john larkin wrote:
    On Tue, 10 Nov 2026 18:38:47 +0800, Johann 'Myrkraverk' Oskarsson >>>>>>> <johann@myrkraverk.invalid> wrote:

    On 11/10/2026 6:12 PM, Liz Tuddenham wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:

    <snip>

    The point of all of this?-a Comedy is fun, you should try it.-a It's
    merely a matter of practice.

    Comedy can be fun, and I've done a bit of it. Claiming that it is just
    a matter of practice means that you don't know what you are talking
    about, as usual.

    <snip>

    Dear Bill,

    Now you're so wrong you have egg on your face.-a How does it feel?

    Like most of the sensations that you imagine I ought to be feeling, I
    don't have any conviction that I've got egg on my face, but I do a
    strong conviction that you are barking up the wrong tree yet again.

    Trying to claim comedy isn't about practice is just your coping mechanism
    for being so unfunny in life everyone stopped talking to you.

    Not a situation that I recognise. Professional comedy is about getting
    the comic timing exactly right, once you've found the comic lines to
    deliver, but that hasn't got much to do with everyday life. Finding
    comic moments in everyday life is a rather different skill.

    So don't project your coping mechanism on other people, it's not
    healthy.

    That might be good advice for people who share your personality defects -
    quite what drives you to pontificate about stuff you know very little
    about isn't all that obvious - but it doesn't generalise to the rest
    of us.
    --
    Bill Sloman, Sydney

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Sat Nov 14 07:33:09 2026
    From Newsgroup: sci.electronics.design

    On 11/10/2026 8:41 PM, Bill Sloman wrote:
    On 10/11/2026 9:20 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 1:49 PM, Bill Sloman wrote:
    On 10/11/2026 2:12 am, john larkin wrote:
    On Mon, 9 Nov 2026 16:42:08 +1100, Bill Sloman <bill.sloman@ieee.org>
    wrote:

    On 9/11/2026 6:27 am, Waldek Hebisch wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:
    On 7/11/2026 3:57 pm, john larkin wrote:
    On Sat, 7 Nov 2026 15:38:13 +1100, Bill Sloman
    <bill.sloman@ieee.org>
    wrote:

    On 7/11/2026 5:54 am, john larkin wrote:
    On Fri, 6 Nov 2026 10:24:07 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and data. >>>>>>>>>>>>>> Signed drivers

    Not a Lisp fan then?

    People love to invent weird computer languages. There are >>>>>>>>>>>> over 3000
    and they keep coming.

    Real men code in binary using the front panel switches.

    I actually used a PDP-7 where you really did have to
    bootstrap it that
    way to read in a paper tape. It was a fun machine as it had a >>>>>>>>>>> vector
    graphics display so was very good for asteroids, but cheated >>>>>>>>>>> at chess.

    I did a lot of stuff in PDP-11 and 68K assembly languages. >>>>>>>>>> With a few
    macros, they are almost higher-level languages.

    All high level languages get compiled into assembly language >>>>>>>>> before they
    can be run on a computer.

    Well, no.

    Counter-example?

    Being a compiler is tedious.

    Unlike some architectures that I might name.

    A language isn't an architecture. It has to be processed into >>>>>>>>> something
    that can be run on hardware (which will have an architecture - >>>>>>>>> it is
    constructed from components that have been connected by some >>>>>>>>> kind of
    architect).

    If that post were compiled to assembly language, it would be a nop. >>>>>>>
    As a compiler, you do seem to be sub-optimal, if not downright
    afunctional. Ignoring implications you don't like isn't an option >>>>>>> for a
    compiler. There can be functions they can't parse - in your case >>>>>>> don't
    want to parse - but that doesn't generate a nop.

    You clearly go into an area where you are not an expert.

    I'm not pretending to be an expert, just less of a clueless newbie in >>>>> the area that John Larkin has shown himself to be.

    I've written several compilers and several RTOS's and a bunch of
    system simulators and hundreds of application programs, a good chunk
    of a million lines of code. From PDP8 to ARM and one CPU that I
    designed myself.

    And I'm sure that you are perfectly happy with all of them.
    Egomaniacs do go around describing themself as "stable geniuses".
    What's your definition of newbie?

    Somebody who doesn't know enough about what he his doing to be aware
    of the areas he doesn't know enough about. It's not so much about
    what you know, as what you know about the limits of your knowledge.

    <snipped pretentious twaddle>


    What? For once I'm trying to have a normal conversation with you, and
    you pull this asinine stunt?

    You
    punk!

    Let me tell you something, punk! You are not the first punk! I've
    dealt with in life, real life, and online; you punk!

    You started to talk about the limits of knowledge, and I added the ob-
    vious corollary that sometimes even you don't know what you don't know.

    Phrased slightly more politely, you punk!

    It's not my fault you're too lazy to look in a dictionary, you punk!

    https://en.wiktionary.org/wiki/unknown_unknown

    https://en.wikipedia.org/wiki/Black_swan_theory

    So now I have done so for you, you punk!


    I don't know why you always think you know best about everything, you
    punk!
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Sat Nov 14 07:40:04 2026
    From Newsgroup: sci.electronics.design

    On 11/12/2026 10:16 AM, Bill Sloman wrote:
    On 11/11/2026 4:35 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/11/2026 12:21 PM, Bill Sloman wrote:
    On 11/11/2026 12:30 am, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 8:46 PM, Bill Sloman wrote:
    On 10/11/2026 9:22 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 6:20 PM, Johann 'Myrkraverk' Oskarsson wrote:

    <snip>

    Best wishes, and keep being miserable for the next fifty years -- at
    least!

    My statistical expectation of survival is rather shorter than that.
    I'd have to become ten years older than longest surviving human ever
    has so far to get there.

    Don't let the statistical expectation stop you.

    Statistical expectation isn't any kind of barrier. It's just a
    prediction based on what has happened in the past.

    Modern medical science does keep people alive for longer than they used
    to manage to survive, but 133 would be a very ripe old age.


    I mean, /I want you/ to
    be here fifty years from now, and still posting your miserable drivel on
    sci.electronics.design, to the delight of the lurkers who simply enjoy
    the consumption of popcorn and coca-cola.-a A good tradition among movie
    goers going back decades, if not centuries.

    What you perceive as miserable drivel may not look quite so miserable
    and drivel-like to better-informed observers (who seem to be fairly numerous).

    And I really don't care what you want. The peanut gallery isn't a discriminating audience.


    What? You want to check out early because you're too lazy to live for
    fifty more years? I mean, sure, I don't control your life, but still
    you've been here for thirty years, so what's fifty more to you?

    And after fifty more years, perhaps you want to be here for another
    fifty years. Don't stop posting on my account. That's just silly.
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Sat Nov 14 07:44:45 2026
    From Newsgroup: sci.electronics.design

    On 11/13/2026 1:35 PM, Bill Sloman wrote:
    On 13/11/2026 9:19 am, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/12/2026 10:31 PM, Bill Sloman wrote:
    In any case, trying to use /knowledge/ as a status symbol seems quite >>>>> stupid to me; yet a lot of people behave this way.

    Correcting errors isn't "using knowledge as a status symbol", except
    in the sense that being able to correct errors does involve asserting
    a superior status, but since the alternative is to leave errors
    unremarked and uncorrected, it is also taking responsibility.

    Of course you think so, but leaving errors is fine too.

    You may think so but you aren't any kind of exemplary character.

    -aAfter all, I'm not being paid to educate you on the errors of your ways.

    What you post isn't exactly educational. It's sort of useful as an
    example of how silly you can get if you don't know what you are talking about, but we aren't short of those kinds of posts here.


    What you are doing now is demonstrating to all of sci.electronics.design
    how stupid and ignorant of the world you are. You are facing the conse- quences of your own actions.

    You started throwing insults in my direction, and you're incapable of apologizing, you punk!

    So now I'm making fun of you at every opportunity, and amusement for my
    own purposes. You are just a source of amusement, no matter what you
    try to claim. You're nothing but a punk who thinks he's important.

    I don't respect you, and I don't even despise you. You're not important enough!


    The rest of sci.electronics.design will just have to sit on the side-
    lines for this, and eat popcorn. I hope they don't run out.
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From JM@sunaecoNoChoppedPork@gmail.com to sci.electronics.design on Sat Nov 14 02:04:38 2026
    From Newsgroup: sci.electronics.design

    On Sat, 14 Nov 2026 07:44:45 +0800, Johann 'Myrkraverk' Oskarsson <johann@myrkraverk.invalid> wrote:

    Bollocks snipped..

    The rest of sci.electronics.design will just have to sit on the side-
    lines for this, and eat popcorn. I hope they don't run out.

    Can't you just go and play somewhere else?

    Perhaps come back when (or if) you grow up?
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Phil Hobbs@pcdhSpamMeSenseless@electrooptical.net to sci.electronics.design on Sat Nov 14 02:40:15 2026
    From Newsgroup: sci.electronics.design

    JM <sunaecoNoChoppedPork@gmail.com> wrote:
    On Sat, 14 Nov 2026 07:44:45 +0800, Johann 'Myrkraverk' Oskarsson <johann@myrkraverk.invalid> wrote:

    Bollocks snipped..

    The rest of sci.electronics.design will just have to sit on the side-
    lines for this, and eat popcorn. I hope they don't run out.

    Can't you just go and play somewhere else?

    Perhaps come back when (or if) you grow up?


    Killfiles are wonderful things.

    Cheers

    Phil Hobbs
    --
    Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC / Hobbs ElectroOptics Optics, Electro-optics, Photonics, Analog Electronics
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Sat Nov 14 14:20:59 2026
    From Newsgroup: sci.electronics.design

    On 11/14/2026 10:04 AM, JM wrote:
    On Sat, 14 Nov 2026 07:44:45 +0800, Johann 'Myrkraverk' Oskarsson <johann@myrkraverk.invalid> wrote:

    Bollocks snipped..

    The rest of sci.electronics.design will just have to sit on the side-
    lines for this, and eat popcorn. I hope they don't run out.

    Can't you just go and play somewhere else?

    Perhaps come back when (or if) you grow up?

    What is this? The in-group vs. the newcomer? Why are you antagonistic
    too?

    And why do you think Bill Sloman shouldn't face the consequences of his
    own action? Is he somehow an anointed priest of sci.electronics.design?
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Sat Nov 14 17:54:16 2026
    From Newsgroup: sci.electronics.design

    On 14/11/2026 10:40 am, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/12/2026 10:16 AM, Bill Sloman wrote:
    On 11/11/2026 4:35 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/11/2026 12:21 PM, Bill Sloman wrote:
    On 11/11/2026 12:30 am, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 8:46 PM, Bill Sloman wrote:
    On 10/11/2026 9:22 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 6:20 PM, Johann 'Myrkraverk' Oskarsson wrote:

    <snip>

    Best wishes, and keep being miserable for the next fifty years -- at >>>>> least!

    My statistical expectation of survival is rather shorter than that.
    I'd have to become ten years older than longest surviving human ever
    has so far to get there.

    Don't let the statistical expectation stop you.

    Statistical expectation isn't any kind of barrier. It's just a
    prediction based on what has happened in the past.

    Modern medical science does keep people alive for longer than they
    used to manage to survive, but 133 would be a very ripe old age.


    I mean, /I want you/ to
    be here fifty years from now, and still posting your miserable drivel on >>> sci.electronics.design, to the delight of the lurkers who simply enjoy
    the consumption of popcorn and coca-cola.-a A good tradition among movie >>> goers going back decades, if not centuries.

    What you perceive as miserable drivel may not look quite so miserable
    and drivel-like to better-informed observers (who seem to be fairly
    numerous).

    And I really don't care what you want. The peanut gallery isn't a
    discriminating audience.


    What?-a You want to check out early because you're too lazy to live for
    fifty more years?-a I mean, sure, I don't control your life, but still
    you've been here for thirty years, so what's fifty more to you?

    Staying alive as you get older takes progressively more effort. I used
    to be able walk for miles, but now I find it difficult to cover the
    half-mile to the local hospital. I've done it twice this week - once to
    get the blood samples taken that my urologist wants to have before he
    talk s to me in a fortnight and once again to take back the 24-hour
    urine sample that he wants analysed.

    And after fifty more years, perhaps you want to be here for another
    fifty years.-a Don't stop posting on my account.-a That's just silly.

    Responding to you at all is totally silly. But it keeps me in practice
    for more worthwhile targets.
    --
    Bill Sloman, SAydney

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Sat Nov 14 18:13:45 2026
    From Newsgroup: sci.electronics.design

    On 14/11/2026 10:33 am, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 8:41 PM, Bill Sloman wrote:
    On 10/11/2026 9:20 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 1:49 PM, Bill Sloman wrote:
    On 10/11/2026 2:12 am, john larkin wrote:
    On Mon, 9 Nov 2026 16:42:08 +1100, Bill Sloman <bill.sloman@ieee.org> >>>>> wrote:

    On 9/11/2026 6:27 am, Waldek Hebisch wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:
    On 7/11/2026 3:57 pm, john larkin wrote:
    On Sat, 7 Nov 2026 15:38:13 +1100, Bill Sloman
    <bill.sloman@ieee.org>
    wrote:

    On 7/11/2026 5:54 am, john larkin wrote:
    On Fri, 6 Nov 2026 10:24:07 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and data. >>>>>>>>>>>>>>> Signed drivers

    Not a Lisp fan then?

    People love to invent weird computer languages. There are >>>>>>>>>>>>> over 3000
    and they keep coming.

    Real men code in binary using the front panel switches. >>>>>>>>>>>>
    I actually used a PDP-7 where you really did have to
    bootstrap it that
    way to read in a paper tape. It was a fun machine as it had >>>>>>>>>>>> a vector
    graphics display so was very good for asteroids, but cheated >>>>>>>>>>>> at chess.

    I did a lot of stuff in PDP-11 and 68K assembly languages. >>>>>>>>>>> With a few
    macros, they are almost higher-level languages.

    All high level languages get compiled into assembly language >>>>>>>>>> before they
    can be run on a computer.

    Well, no.

    Counter-example?

    Being a compiler is tedious.

    Unlike some architectures that I might name.

    A language isn't an architecture. It has to be processed into >>>>>>>>>> something
    that can be run on hardware (which will have an architecture - >>>>>>>>>> it is
    constructed from components that have been connected by some >>>>>>>>>> kind of
    architect).

    If that post were compiled to assembly language, it would be a >>>>>>>>> nop.

    As a compiler, you do seem to be sub-optimal, if not downright >>>>>>>> afunctional. Ignoring implications you don't like isn't an
    option for a
    compiler. There can be functions they can't parse - in your case >>>>>>>> don't
    want to parse - but that doesn't generate a nop.

    You clearly go into an area where you are not an expert.

    I'm not pretending to be an expert, just less of a clueless newbie in >>>>>> the area that John Larkin has shown himself to be.

    I've written several compilers and several RTOS's and a bunch of
    system simulators and hundreds of application programs, a good chunk >>>>> of a million lines of code. From PDP8 to ARM and one CPU that I
    designed myself.

    And I'm sure that you are perfectly happy with all of them.
    Egomaniacs do go around describing themself as "stable geniuses".
    What's your definition of newbie?

    Somebody who doesn't know enough about what he his doing to be aware
    of the areas he doesn't know enough about. It's not so much about
    what you know, as what you know about the limits of your knowledge.

    <snipped pretentious twaddle>


    What?-a For once I'm trying to have a normal conversation with you, and
    you pull this asinine stunt?

    -a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a You
    -a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a punk!

    Let me tell you something, punk!-a You are not the first punk!-a I've
    dealt with in life, real life, and online; you punk!

    You started to talk about the limits of knowledge, and I added the
    obvious corollary that sometimes even you don't know what you don't know.

    Phrased slightly more politely, you punk!

    It's not my fault you're too lazy to look in a dictionary, you punk!

    I mostly don't need to. I read lot when young, and do have a good memory.

    -a https://en.wiktionary.org/wiki/unknown_unknown

    -a https://en.wikipedia.org/wiki/Black_swan_theory

    So now I have done so for you, you punk!

    You didn't need to bother

    /wiki/The_Black_Swan:_The_Impact_of_the_Highly_Improbable

    I bought the book when Nassim Nicholas Taleb first published it in
    2007and read it from cover to cover, which wasn't difficult.
    It's sort of interesting.

    I don't know why you always think you know best about everything, you
    punk!

    I don't. I frequently know better than the self-important twits who post
    here from time to time, but I have had a good enough education to know
    how little I know. I read enough to learn more as it gets discovered and publicised, and I know enough to be well aware that a lot of interesting
    stuff isn't interesting to enough people to get publicised.

    The punk here is you. You wouldn't waste your time boasting about
    "dealing with me" if you had any grasp of reality.

    Donald J. Trump boasts about writing "The Art of the Deal" when it is
    well known that it was ghost-written for him. You seem to have the same
    kind of delusions about what "dealing with" actually implies.
    --
    Bill Sloman, Sydney

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Sat Nov 14 18:20:08 2026
    From Newsgroup: sci.electronics.design

    On 14/11/2026 5:20 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/14/2026 10:04 AM, JM wrote:
    On Sat, 14 Nov 2026 07:44:45 +0800, Johann 'Myrkraverk' Oskarsson
    <johann@myrkraverk.invalid> wrote:

    Bollocks snipped..

    The rest of sci.electronics.design will just have to sit on the side-
    lines for this, and eat popcorn.-a I hope they don't run out.

    Can't you just go and play somewhere else?

    Perhaps come back when (or if) you grow up?

    What is this?-a The in-group vs. the newcomer?-a Why are you antagonistic too?

    And why do you think Bill Sloman shouldn't face the consequences of his
    own action?-a Is he somehow an anointed priest of sci.electronics.design?

    You do seem to think that your own actions are consquential. JM and Phil
    Hobbs have been around here long enough to have seen any number of
    voluble half-wits run out of steam.
    --
    Bill Sloman, Sydney
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Sat Nov 14 19:54:41 2026
    From Newsgroup: sci.electronics.design

    On 11/14/2026 3:20 PM, Bill Sloman wrote:
    On 14/11/2026 5:20 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/14/2026 10:04 AM, JM wrote:
    On Sat, 14 Nov 2026 07:44:45 +0800, Johann 'Myrkraverk' Oskarsson
    <johann@myrkraverk.invalid> wrote:

    Bollocks snipped..

    The rest of sci.electronics.design will just have to sit on the side-
    lines for this, and eat popcorn.-a I hope they don't run out.

    Can't you just go and play somewhere else?

    Perhaps come back when (or if) you grow up?

    What is this?-a The in-group vs. the newcomer?-a Why are you antagonistic
    too?

    And why do you think Bill Sloman shouldn't face the consequences of his
    own action?-a Is he somehow an anointed priest of sci.electronics.design?

    You do seem to think that your own actions are consquential. JM and Phil Hobbs have been around here long enough to have seen any number of
    voluble half-wits run out of steam.


    You have nothing to worry about. I deleted my Steam account, so I won't
    run out of Steam! I already did.
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Sat Nov 14 19:56:12 2026
    From Newsgroup: sci.electronics.design

    On 11/14/2026 2:54 PM, Bill Sloman wrote:
    On 14/11/2026 10:40 am, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/12/2026 10:16 AM, Bill Sloman wrote:
    On 11/11/2026 4:35 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/11/2026 12:21 PM, Bill Sloman wrote:
    On 11/11/2026 12:30 am, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 8:46 PM, Bill Sloman wrote:
    On 10/11/2026 9:22 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 6:20 PM, Johann 'Myrkraverk' Oskarsson wrote:

    <snip>

    Best wishes, and keep being miserable for the next fifty years -- at >>>>>> least!

    My statistical expectation of survival is rather shorter than that. >>>>> I'd have to become ten years older than longest surviving human
    ever has so far to get there.

    Don't let the statistical expectation stop you.

    Statistical expectation isn't any kind of barrier. It's just a
    prediction based on what has happened in the past.

    Modern medical science does keep people alive for longer than they
    used to manage to survive, but 133 would be a very ripe old age.


    I mean, /I want you/ to
    be here fifty years from now, and still posting your miserable
    drivel on
    sci.electronics.design, to the delight of the lurkers who simply enjoy >>>> the consumption of popcorn and coca-cola.-a A good tradition among movie >>>> goers going back decades, if not centuries.

    What you perceive as miserable drivel may not look quite so miserable
    and drivel-like to better-informed observers (who seem to be fairly
    numerous).

    And I really don't care what you want. The peanut gallery isn't a
    discriminating audience.


    What?-a You want to check out early because you're too lazy to live for
    fifty more years?-a I mean, sure, I don't control your life, but still
    you've been here for thirty years, so what's fifty more to you?

    Staying alive as you get older takes progressively more effort. I used
    to be able walk for miles, but now I find it difficult to cover the half-mile to the local hospital. I've done it twice this week - once to
    get the blood samples taken that my urologist wants to have before he
    talk s to me in a-a fortnight and once again to take back the 24-hour
    urine sample that he wants analysed.

    And after fifty more years, perhaps you want to be here for another
    fifty years.-a Don't stop posting on my account.-a That's just silly.

    Responding to you at all is totally silly. But it keeps me in practice
    for more worthwhile targets.


    It's good that you're staying in practice. Both at staying alive, and responding to me. I have no problems with either.
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Sat Nov 14 20:06:35 2026
    From Newsgroup: sci.electronics.design

    On 11/14/2026 3:13 PM, Bill Sloman wrote:
    On 14/11/2026 10:33 am, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 8:41 PM, Bill Sloman wrote:
    On 10/11/2026 9:20 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 1:49 PM, Bill Sloman wrote:
    On 10/11/2026 2:12 am, john larkin wrote:
    On Mon, 9 Nov 2026 16:42:08 +1100, Bill Sloman <bill.sloman@ieee.org> >>>>>> wrote:

    On 9/11/2026 6:27 am, Waldek Hebisch wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:
    On 7/11/2026 3:57 pm, john larkin wrote:
    On Sat, 7 Nov 2026 15:38:13 +1100, Bill Sloman
    <bill.sloman@ieee.org>
    wrote:

    On 7/11/2026 5:54 am, john larkin wrote:
    On Fri, 6 Nov 2026 10:24:07 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown >>>>>>>>>>>>>>> <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and >>>>>>>>>>>>>>>> data. Signed drivers

    Not a Lisp fan then?

    People love to invent weird computer languages. There are >>>>>>>>>>>>>> over 3000
    and they keep coming.

    Real men code in binary using the front panel switches. >>>>>>>>>>>>>
    I actually used a PDP-7 where you really did have to >>>>>>>>>>>>> bootstrap it that
    way to read in a paper tape. It was a fun machine as it had >>>>>>>>>>>>> a vector
    graphics display so was very good for asteroids, but >>>>>>>>>>>>> cheated at chess.

    I did a lot of stuff in PDP-11 and 68K assembly languages. >>>>>>>>>>>> With a few
    macros, they are almost higher-level languages.

    All high level languages get compiled into assembly language >>>>>>>>>>> before they
    can be run on a computer.

    Well, no.

    Counter-example?

    Being a compiler is tedious.

    Unlike some architectures that I might name.

    A language isn't an architecture. It has to be processed into >>>>>>>>>>> something
    that can be run on hardware (which will have an architecture >>>>>>>>>>> - it is
    constructed from components that have been connected by some >>>>>>>>>>> kind of
    architect).

    If that post were compiled to assembly language, it would be a >>>>>>>>>> nop.

    As a compiler, you do seem to be sub-optimal, if not downright >>>>>>>>> afunctional. Ignoring implications you don't like isn't an
    option for a
    compiler. There can be functions they can't parse - in your >>>>>>>>> case don't
    want to parse - but that doesn't generate a nop.

    You clearly go into an area where you are not an expert.

    I'm not pretending to be an expert, just less of a clueless
    newbie in
    the area that John Larkin has shown himself to be.

    I've written several compilers and several RTOS's and a bunch of
    system simulators and hundreds of application programs, a good chunk >>>>>> of a million lines of code. From PDP8 to ARM and one CPU that I
    designed myself.

    And I'm sure that you are perfectly happy with all of them.
    Egomaniacs do go around describing themself as "stable geniuses".
    What's your definition of newbie?

    Somebody who doesn't know enough about what he his doing to be
    aware of the areas he doesn't know enough about. It's not so much
    about what you know, as what you know about the limits of your
    knowledge.

    <snipped pretentious twaddle>


    What?-a For once I'm trying to have a normal conversation with you, and
    you pull this asinine stunt?

    -a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a You
    -a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a punk!

    Let me tell you something, punk!-a You are not the first punk!-a I've
    dealt with in life, real life, and online; you punk!

    You started to talk about the limits of knowledge, and I added the
    obvious corollary that sometimes even you don't know what you don't know.

    Phrased slightly more politely, you punk!

    It's not my fault you're too lazy to look in a dictionary, you punk!

    I mostly don't need to. I read lot when young, and do have a good memory.

    -a-a https://en.wiktionary.org/wiki/unknown_unknown

    -a-a https://en.wikipedia.org/wiki/Black_swan_theory

    So now I have done so for you, you punk!

    You didn't need to bother

    /wiki/The_Black_Swan:_The_Impact_of_the_Highly_Improbable

    I bought the book when-a Nassim Nicholas Taleb first published it in
    2007and read it from cover to cover, which wasn't difficult.
    It's sort of interesting.

    I don't know why you always think you know best about everything, you
    punk!

    I don't. I frequently know better than the self-important twits who post here from time to time, but I have had a good enough education to know
    how little I know. I read enough to learn more as it gets discovered and publicised, and I know enough to be well aware that a lot of interesting stuff isn't interesting to enough people to get publicised.

    The punk here is you. You wouldn't waste your time boasting about
    "dealing with me" if you had any grasp of reality.

    Donald J. Trump boasts about writing "The Art of the Deal" when it is
    well known that it was ghost-written for him. You seem to have the same
    kind of delusions about what "dealing with" actually implies.


    I don't really understand why you bring him up. I don't know the guy,
    as I've never met him. And I haven't even read his book, whoever wrote
    it.

    A much more interesting book is /The End of Alchemy/ (2017) by Mervyn
    King. It is not about alchemy, nor chemistry [1] and the byline gives
    it away, /Money, Banking, and the Future of Global Economy/.

    Mervyn King was a member of the Royal Mint I believe, but it's been a
    while since I read it.

    Now, since you've read the book /The Black Swan/, how many black swan
    events have you caused? Or do you think there's no correlation between
    reading a book, and causing an event? Can Karl Marx say the same [2]?


    Anyway, don't stop on my account, you can still cause a few black swan
    events in your life, if you're not lazy.


    Best wishes, and happy swans!

    [1] Which could be interesting enough, given your background in chem-
    istry.

    [2] Well, technically, it's /writing a book/ for him, but that's just
    a minor difference, isn't it?
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Sun Nov 15 01:14:59 2026
    From Newsgroup: sci.electronics.design

    On 14/11/2026 10:54 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/14/2026 3:20 PM, Bill Sloman wrote:
    On 14/11/2026 5:20 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/14/2026 10:04 AM, JM wrote:
    On Sat, 14 Nov 2026 07:44:45 +0800, Johann 'Myrkraverk' Oskarsson
    <johann@myrkraverk.invalid> wrote:

    Bollocks snipped..

    The rest of sci.electronics.design will just have to sit on the side- >>>>> lines for this, and eat popcorn.-a I hope they don't run out.

    Can't you just go and play somewhere else?

    Perhaps come back when (or if) you grow up?

    What is this?-a The in-group vs. the newcomer?-a Why are you antagonistic >>> too?

    And why do you think Bill Sloman shouldn't face the consequences of his
    own action?-a Is he somehow an anointed priest of sci.electronics.design? >>
    You do seem to think that your own actions are consequential. JM and
    Phil Hobbs have been around here long enough to have seen any number
    of voluble half-wits run out of steam.

    You have nothing to worry about.-a I deleted my Steam account, so I won't
    run out of Steam!-a I already did.

    The phrase is rather older than computer gaming, and refers to steam
    engines. Think George Stevensons's Rocket of 1829.

    https://en.wikipedia.org/wiki/Stephenson%27s_Rocket
    --
    Bill Sloman, Sydney
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Sun Nov 15 02:12:32 2026
    From Newsgroup: sci.electronics.design

    On 14/11/2026 10:56 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/14/2026 2:54 PM, Bill Sloman wrote:
    On 14/11/2026 10:40 am, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/12/2026 10:16 AM, Bill Sloman wrote:
    On 11/11/2026 4:35 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/11/2026 12:21 PM, Bill Sloman wrote:
    On 11/11/2026 12:30 am, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 8:46 PM, Bill Sloman wrote:
    On 10/11/2026 9:22 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 6:20 PM, Johann 'Myrkraverk' Oskarsson wrote:

    <snip>

    Best wishes, and keep being miserable for the next fifty years -- at >>>>>>> least!

    My statistical expectation of survival is rather shorter than
    that. I'd have to become ten years older than longest surviving
    human ever has so far to get there.

    Don't let the statistical expectation stop you.

    Statistical expectation isn't any kind of barrier. It's just a
    prediction based on what has happened in the past.

    Modern medical science does keep people alive for longer than they
    used to manage to survive, but 133 would be a very ripe old age.


    I mean, /I want you/ to
    be here fifty years from now, and still posting your miserable
    drivel on
    sci.electronics.design, to the delight of the lurkers who simply enjoy >>>>> the consumption of popcorn and coca-cola.-a A good tradition among
    movie
    goers going back decades, if not centuries.

    What you perceive as miserable drivel may not look quite so
    miserable and drivel-like to better-informed observers (who seem to
    be fairly numerous).

    And I really don't care what you want. The peanut gallery isn't a
    discriminating audience.


    What?-a You want to check out early because you're too lazy to live for
    fifty more years?-a I mean, sure, I don't control your life, but still
    you've been here for thirty years, so what's fifty more to you?

    Staying alive as you get older takes progressively more effort. I used
    to be able walk for miles, but now I find it difficult to cover the
    half-mile to the local hospital. I've done it twice this week - once
    to get the blood samples taken that my urologist wants to have before
    he talk s to me in a-a fortnight and once again to take back the
    24-hour urine sample that he wants analysed.

    And after fifty more years, perhaps you want to be here for another
    fifty years.-a Don't stop posting on my account.-a That's just silly.

    Responding to you at all is totally silly. But it keeps me in practice
    for more worthwhile targets.


    It's good that you're staying in practice.-a Both at staying alive, and responding to me.-a I have no problems with either.

    If yours were the only responses I was getting, I wouldn't bother.
    --
    Bill Sloman, Sydney
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Sun Nov 15 00:30:08 2026
    From Newsgroup: sci.electronics.design

    On 11/14/2026 10:14 PM, Bill Sloman wrote:
    On 14/11/2026 10:54 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/14/2026 3:20 PM, Bill Sloman wrote:
    On 14/11/2026 5:20 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/14/2026 10:04 AM, JM wrote:
    On Sat, 14 Nov 2026 07:44:45 +0800, Johann 'Myrkraverk' Oskarsson
    <johann@myrkraverk.invalid> wrote:

    Bollocks snipped..

    The rest of sci.electronics.design will just have to sit on the side- >>>>>> lines for this, and eat popcorn.-a I hope they don't run out.

    Can't you just go and play somewhere else?

    Perhaps come back when (or if) you grow up?

    What is this?-a The in-group vs. the newcomer?-a Why are you antagonistic >>>> too?

    And why do you think Bill Sloman shouldn't face the consequences of his >>>> own action?-a Is he somehow an anointed priest of
    sci.electronics.design?

    You do seem to think that your own actions are consequential. JM and
    Phil Hobbs have been around here long enough to have seen any number
    of voluble half-wits run out of steam.

    You have nothing to worry about.-a I deleted my Steam account, so I won't
    run out of Steam!-a I already did.

    The phrase is rather older than computer gaming, and refers to steam engines. Think George Stevensons's Rocket of 1829.

    https://en.wikipedia.org/wiki/Stephenson%27s_Rocket


    Yes, I know that. I just don't care.
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Sun Nov 15 02:53:57 2026
    From Newsgroup: sci.electronics.design

    On 14/11/2026 11:06 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/14/2026 3:13 PM, Bill Sloman wrote:
    On 14/11/2026 10:33 am, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 8:41 PM, Bill Sloman wrote:
    On 10/11/2026 9:20 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 1:49 PM, Bill Sloman wrote:
    On 10/11/2026 2:12 am, john larkin wrote:
    On Mon, 9 Nov 2026 16:42:08 +1100, Bill Sloman
    <bill.sloman@ieee.org>
    wrote:

    On 9/11/2026 6:27 am, Waldek Hebisch wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:
    On 7/11/2026 3:57 pm, john larkin wrote:
    On Sat, 7 Nov 2026 15:38:13 +1100, Bill Sloman
    <bill.sloman@ieee.org>
    wrote:

    On 7/11/2026 5:54 am, john larkin wrote:
    On Fri, 6 Nov 2026 10:24:07 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown >>>>>>>>>>>>>>>> <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and >>>>>>>>>>>>>>>>> data. Signed drivers

    Not a Lisp fan then?

    People love to invent weird computer languages. There are >>>>>>>>>>>>>>> over 3000
    and they keep coming.

    Real men code in binary using the front panel switches. >>>>>>>>>>>>>>
    I actually used a PDP-7 where you really did have to >>>>>>>>>>>>>> bootstrap it that
    way to read in a paper tape. It was a fun machine as it >>>>>>>>>>>>>> had a vector
    graphics display so was very good for asteroids, but >>>>>>>>>>>>>> cheated at chess.

    I did a lot of stuff in PDP-11 and 68K assembly languages. >>>>>>>>>>>>> With a few
    macros, they are almost higher-level languages.

    All high level languages get compiled into assembly language >>>>>>>>>>>> before they
    can be run on a computer.

    Well, no.

    Counter-example?

    Being a compiler is tedious.

    Unlike some architectures that I might name.

    A language isn't an architecture. It has to be processed >>>>>>>>>>>> into something
    that can be run on hardware (which will have an architecture >>>>>>>>>>>> - it is
    constructed from components that have been connected by some >>>>>>>>>>>> kind of
    architect).

    If that post were compiled to assembly language, it would be >>>>>>>>>>> a nop.

    As a compiler, you do seem to be sub-optimal, if not downright >>>>>>>>>> afunctional. Ignoring implications you don't like isn't an >>>>>>>>>> option for a
    compiler. There can be functions they can't parse - in your >>>>>>>>>> case don't
    want to parse - but that doesn't generate a nop.

    You clearly go into an area where you are not an expert.

    I'm not pretending to be an expert, just less of a clueless
    newbie in
    the area that John Larkin has shown himself to be.

    I've written several compilers and several RTOS's and a bunch of >>>>>>> system simulators and hundreds of application programs, a good chunk >>>>>>> of a million lines of code. From PDP8 to ARM and one CPU that I
    designed myself.

    And I'm sure that you are perfectly happy with all of them.
    Egomaniacs do go around describing themself as "stable geniuses". >>>>>>> What's your definition of newbie?

    Somebody who doesn't know enough about what he his doing to be
    aware of the areas he doesn't know enough about. It's not so much >>>>>> about what you know, as what you know about the limits of your
    knowledge.

    <snipped pretentious twaddle>


    What?-a For once I'm trying to have a normal conversation with you, and
    you pull this asinine stunt?

    -a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a You
    -a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a punk! >>>
    Let me tell you something, punk!-a You are not the first punk!-a I've
    dealt with in life, real life, and online; you punk!

    You started to talk about the limits of knowledge, and I added the
    obvious corollary that sometimes even you don't know what you don't
    know.

    Phrased slightly more politely, you punk!

    It's not my fault you're too lazy to look in a dictionary, you punk!

    I mostly don't need to. I read lot when young, and do have a good memory.

    -a-a https://en.wiktionary.org/wiki/unknown_unknown

    -a-a https://en.wikipedia.org/wiki/Black_swan_theory

    So now I have done so for you, you punk!

    You didn't need to bother

    /wiki/The_Black_Swan:_The_Impact_of_the_Highly_Improbable

    I bought the book when-a Nassim Nicholas Taleb first published it in
    2007and read it from cover to cover, which wasn't difficult.
    It's sort of interesting.

    I don't know why you always think you know best about everything, you
    punk!

    I don't. I frequently know better than the self-important twits who
    post here from time to time, but I have had a good enough education to
    know how little I know. I read enough to learn more as it gets
    discovered and publicised, and I know enough to be well aware that a
    lot of interesting stuff isn't interesting to enough people to get
    publicised.

    The punk here is you. You wouldn't waste your time boasting about
    "dealing with me" if you had any grasp of reality.

    Donald J. Trump boasts about writing "The Art of the Deal" when it is
    well known that it was ghost-written for him. You seem to have the
    same kind of delusions about what "dealing with" actually implies.

    I don't really understand why you bring him up.-a I don't know the guy,
    as I've never met him.-a And I haven't even read his book, whoever wrote
    it.

    But you do share one of his intellectual weaknesses.

    A much more interesting book is /The End of Alchemy/ (2017) by Mervyn
    King.-a It is not about alchemy, nor chemistry [1] and the byline gives
    it away, /Money, Banking, and the Future of Global Economy/.

    Mervyn King was a member of the Royal Mint I believe, but it's been a
    while since I read it.

    Now, since you've read the book /The Black Swan/, how many black swan
    events have you caused?

    I've got three patents, which probably do qualify as black swan events -
    it the ideas had been predictable they would have been obvious to those skilled in the art.

    Or do you think there's no correlation between
    reading a book, and causing an event?-a Can Karl Marx say the same [2]?

    Jonathon Israel thinks that Spinoza writing a couple of books had a lot
    to do with the democratic revolution.

    https://www.jstor.org/stable/j.ctt7t0f8

    he's written more and longer books on that and related subjects. I got
    onto him when I moved to the Netherlands and read his history of the
    Dutch Republic.

    Books reflect the issues of the day, and can influence them. Karl Marx
    was a remarkably influential figure, and his ideas about "the leading
    role of the party" were unfortunately influential. Much of what he had
    to say was very sensible, but that lot wasn't.

    Anyway, don't stop on my account, you can still cause a few black swan
    events in your life, if you're not lazy.

    Laziness probably doesn't come into it. Black swan events are
    unpredictable, and if you can't predict what effect unexpected idea will
    have you can't put a lot of energy into encouraging the effects.
    [1] Which could be interesting enough, given your background in
    chemistry.

    It wasn't just my background. Both my parents had university degrees in chemistry, but nether was remotely interested in alchemy.

    [2] Well, technically, it's /writing a book/ for him, but that's just
    a minor difference, isn't it?

    The guy who ghost-wrote "The Art of the Deal" for Donald Trump had to
    put a lot of effort making what he wrote look as if Donald Trump might
    have written it, and he complained afterwards that it was difficult to
    get enough out of Donald Trump on the subject to create a document that
    might look as if Donald Trump had written it. A ninety-second attention
    span doesn't generate a lot of text in one hit, so the ghost-writer had
    to glom together a lot of sound-bites.
    --
    Bill Sloman, Sydney

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Johann 'Myrkraverk' Oskarsson@johann@myrkraverk.invalid to sci.electronics.design on Sun Nov 15 01:20:07 2026
    From Newsgroup: sci.electronics.design

    On 11/14/2026 11:53 PM, Bill Sloman wrote:
    On 14/11/2026 11:06 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/14/2026 3:13 PM, Bill Sloman wrote:
    On 14/11/2026 10:33 am, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 8:41 PM, Bill Sloman wrote:
    On 10/11/2026 9:20 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 1:49 PM, Bill Sloman wrote:
    On 10/11/2026 2:12 am, john larkin wrote:
    On Mon, 9 Nov 2026 16:42:08 +1100, Bill Sloman
    <bill.sloman@ieee.org>
    wrote:

    On 9/11/2026 6:27 am, Waldek Hebisch wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:
    On 7/11/2026 3:57 pm, john larkin wrote:
    On Sat, 7 Nov 2026 15:38:13 +1100, Bill Sloman
    <bill.sloman@ieee.org>
    wrote:

    On 7/11/2026 5:54 am, john larkin wrote:
    On Fri, 6 Nov 2026 10:24:07 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown >>>>>>>>>>>>>>>>> <'''newspam'''@nonad.co.uk> wrote:

    segmented memory and hard separation of program and >>>>>>>>>>>>>>>>>> data. Signed drivers

    Not a Lisp fan then?

    People love to invent weird computer languages. There >>>>>>>>>>>>>>>> are over 3000
    and they keep coming.

    Real men code in binary using the front panel switches. >>>>>>>>>>>>>>>
    I actually used a PDP-7 where you really did have to >>>>>>>>>>>>>>> bootstrap it that
    way to read in a paper tape. It was a fun machine as it >>>>>>>>>>>>>>> had a vector
    graphics display so was very good for asteroids, but >>>>>>>>>>>>>>> cheated at chess.

    I did a lot of stuff in PDP-11 and 68K assembly languages. >>>>>>>>>>>>>> With a few
    macros, they are almost higher-level languages.

    All high level languages get compiled into assembly >>>>>>>>>>>>> language before they
    can be run on a computer.

    Well, no.

    Counter-example?

    Being a compiler is tedious.

    Unlike some architectures that I might name.

    A language isn't an architecture. It has to be processed >>>>>>>>>>>>> into something
    that can be run on hardware (which will have an
    architecture - it is
    constructed from components that have been connected by >>>>>>>>>>>>> some kind of
    architect).

    If that post were compiled to assembly language, it would be >>>>>>>>>>>> a nop.

    As a compiler, you do seem to be sub-optimal, if not downright >>>>>>>>>>> afunctional. Ignoring implications you don't like isn't an >>>>>>>>>>> option for a
    compiler. There can be functions they can't parse - in your >>>>>>>>>>> case don't
    want to parse - but that doesn't generate a nop.

    You clearly go into an area where you are not an expert.

    I'm not pretending to be an expert, just less of a clueless >>>>>>>>> newbie in
    the area that John Larkin has shown himself to be.

    I've written several compilers and several RTOS's and a bunch of >>>>>>>> system simulators and hundreds of application programs, a good >>>>>>>> chunk
    of a million lines of code. From PDP8 to ARM and one CPU that I >>>>>>>> designed myself.

    And I'm sure that you are perfectly happy with all of them.
    Egomaniacs do go around describing themself as "stable geniuses". >>>>>>>> What's your definition of newbie?

    Somebody who doesn't know enough about what he his doing to be
    aware of the areas he doesn't know enough about. It's not so much >>>>>>> about what you know, as what you know about the limits of your
    knowledge.

    <snipped pretentious twaddle>


    What?-a For once I'm trying to have a normal conversation with you, and >>>> you pull this asinine stunt?

    -a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a You
    -a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a-a punk! >>>>
    Let me tell you something, punk!-a You are not the first punk!-a I've
    dealt with in life, real life, and online; you punk!

    You started to talk about the limits of knowledge, and I added the
    obvious corollary that sometimes even you don't know what you don't
    know.

    Phrased slightly more politely, you punk!

    It's not my fault you're too lazy to look in a dictionary, you punk!

    I mostly don't need to. I read lot when young, and do have a good
    memory.

    -a-a https://en.wiktionary.org/wiki/unknown_unknown

    -a-a https://en.wikipedia.org/wiki/Black_swan_theory

    So now I have done so for you, you punk!

    You didn't need to bother

    /wiki/The_Black_Swan:_The_Impact_of_the_Highly_Improbable

    I bought the book when-a Nassim Nicholas Taleb first published it in
    2007and read it from cover to cover, which wasn't difficult.
    It's sort of interesting.

    I don't know why you always think you know best about everything, you
    punk!

    I don't. I frequently know better than the self-important twits who
    post here from time to time, but I have had a good enough education
    to know how little I know. I read enough to learn more as it gets
    discovered and publicised, and I know enough to be well aware that a
    lot of interesting stuff isn't interesting to enough people to get
    publicised.

    The punk here is you. You wouldn't waste your time boasting about
    "dealing with me" if you had any grasp of reality.

    Donald J. Trump boasts about writing "The Art of the Deal" when it is
    well known that it was ghost-written for him. You seem to have the
    same kind of delusions about what "dealing with" actually implies.

    I don't really understand why you bring him up.-a I don't know the guy,
    as I've never met him.-a And I haven't even read his book, whoever wrote
    it.

    But you do share one of his intellectual weaknesses.


    So do you, so why is that relevant to anything? I mean, the guy's fa-
    mous sure, but that's no reason to bring him into this conversation, un-
    less you want to invite him to post here in sci.electronics.design.

    Now, that would be something.


    A much more interesting book is /The End of Alchemy/ (2017) by Mervyn
    King.-a It is not about alchemy, nor chemistry [1] and the byline gives
    it away, /Money, Banking, and the Future of Global Economy/.

    Mervyn King was a member of the Royal Mint I believe, but it's been a
    while since I read it.

    Now, since you've read the book /The Black Swan/, how many black swan
    events have you caused?

    I've got three patents, which probably do qualify as black swan events -
    it the ideas had been predictable they would have been obvious to those skilled in the art.

    Congratulations-ish on that.


    Or do you think there's no correlation between
    reading a book, and causing an event?-a Can Karl Marx say the same [2]?

    Jonathon Israel thinks that Spinoza writing a couple of books had a lot
    to do with the democratic revolution.

    https://www.jstor.org/stable/j.ctt7t0f8

    I've added this book to my /saved for later/ on Abebooks. I won't do
    more for now.


    he's written more and longer books on that and related subjects. I got
    onto him when I moved to the Netherlands and read his history of the
    Dutch Republic.

    Books reflect the issues of the day, and can influence them. Karl Marx
    was a remarkably influential figure, and his ideas about "the leading
    role of the party" were unfortunately influential. Much of what he had
    to say was very sensible, but that lot wasn't.

    Anyway, don't stop on my account, you can still cause a few black swan
    events in your life, if you're not lazy.

    Laziness probably doesn't come into it. Black swan events are
    unpredictable, and if you can't predict what effect unexpected idea will have you can't put a lot of energy into encouraging the effects.

    [1] Which could be interesting enough, given your background in
    chemistry.

    It wasn't just my background. Both my parents had university degrees in chemistry,-a but nether was remotely interested in alchemy.

    Alchemy as the precursor to modern chemistry is quite fascinating. I
    did read up on Antoine Lavoisier, but only the superficial stuff readily available on the web. I haven't dug into any detailed biographies yet.

    Abu Musa Jabir ibn Hayyan is more famous for bringing us the term /gib-
    berish/ which I'm sure you're well acquainted with.


    [2] Well, technically, it's /writing a book/ for him, but that's just
    a minor difference, isn't it?

    The guy who ghost-wrote "The Art of the Deal" for Donald Trump had to
    put a lot of effort making what he wrote look as if Donald Trump might
    have written it, and he complained afterwards that it was difficult to
    get enough out of Donald Trump on the subject to create a document that might look as if Donald Trump had written it. A ninety-second attention
    span doesn't generate a lot of text in one hit, so the ghost-writer had
    to glom together a lot of sound-bites.


    Well, as I don't know Trump personally, I don't feel any need to bring
    him up in every conversation. I suggest you do the same.

    I'm not interested in /The Art of the Deal/, and not because Trump did
    not write it himself. It would be much more interesting /had/ he writ-
    ten it himself, I admit. I mean, the man must have learned to type at
    some point in life, if only to play /Leather Goddesses of Phobos/.

    I never finished the game myself, but I'm sure I'll get to it /one of
    these years/. Did you play it?


    Best wishes, and happy Z-machine games!
    --
    Johann | email: invalid -> com | http://www.myrkraverk.com/blog/
    I'm not from the Internet, I just work there. | via Easynews.com https://bsky.app/profile/myrkraverk.bsky.social | for ( ;; ) _:;
    Federated at https://fed.brid.gy/bsky/myrkraverk.bsky.social
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Sun Nov 15 15:38:38 2026
    From Newsgroup: sci.electronics.design

    On 15/11/2026 4:20 am, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/14/2026 11:53 PM, Bill Sloman wrote:
    On 14/11/2026 11:06 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/14/2026 3:13 PM, Bill Sloman wrote:
    On 14/11/2026 10:33 am, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 8:41 PM, Bill Sloman wrote:
    On 10/11/2026 9:20 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/10/2026 1:49 PM, Bill Sloman wrote:
    On 10/11/2026 2:12 am, john larkin wrote:
    On Mon, 9 Nov 2026 16:42:08 +1100, Bill Sloman
    <bill.sloman@ieee.org>
    wrote:

    On 9/11/2026 6:27 am, Waldek Hebisch wrote:
    Bill Sloman <bill.sloman@ieee.org> wrote:
    On 7/11/2026 3:57 pm, john larkin wrote:
    On Sat, 7 Nov 2026 15:38:13 +1100, Bill Sloman
    <bill.sloman@ieee.org>
    wrote:

    On 7/11/2026 5:54 am, john larkin wrote:
    On Fri, 6 Nov 2026 10:24:07 +0000, Martin Brown
    <'''newspam'''@nonad.co.uk> wrote:

    On 30/10/2026 16:15, john larkin wrote:
    On Fri, 30 Oct 2026 15:36:05 +0000, JM
    <sunaecoNoChoppedPork@gmail.com> wrote:

    On Thu, 29 Oct 2026 17:36:05 +0000, Martin Brown >>>>>>>>>>>>>>>>>> <'''newspam'''@nonad.co.uk> wrote:

    <snip>

    It's not my fault you're too lazy to look in a dictionary, you punk!

    I mostly don't need to. I read lot when young, and do have a good
    memory.

    -a-a https://en.wiktionary.org/wiki/unknown_unknown

    -a-a https://en.wikipedia.org/wiki/Black_swan_theory

    So now I have done so for you, you punk!

    You didn't need to bother

    /wiki/The_Black_Swan:_The_Impact_of_the_Highly_Improbable

    I bought the book when-a Nassim Nicholas Taleb first published it in
    2007and read it from cover to cover, which wasn't difficult.
    It's sort of interesting.

    I don't know why you always think you know best about everything, you >>>>> punk!

    I don't. I frequently know better than the self-important twits who
    post here from time to time, but I have had a good enough education
    to know how little I know. I read enough to learn more as it gets
    discovered and publicised, and I know enough to be well aware that a
    lot of interesting stuff isn't interesting to enough people to get
    publicised.

    The punk here is you. You wouldn't waste your time boasting about
    "dealing with me" if you had any grasp of reality.

    Donald J. Trump boasts about writing "The Art of the Deal" when it
    is well known that it was ghost-written for him. You seem to have
    the same kind of delusions about what "dealing with" actually implies.

    I don't really understand why you bring him up.-a I don't know the guy,
    as I've never met him.-a And I haven't even read his book, whoever wrote >>> it.

    But you do share one of his intellectual weaknesses.

    So do you,

    Really? Which one?

    so why is that relevant to anything?-a I mean, the guy's famous
    ]sure, but that's no reason to bring him into this conversation,
    unless you want to invite him to post here in sci.electronics.design.

    He's a famous horrible example of people who think that they know how to "deal" with stuff, and haven't got a clue.

    Now, that would be something.

    It would certainly be odd, but Trump does behave oddly.

    A much more interesting book is /The End of Alchemy/ (2017) by Mervyn
    King.-a It is not about alchemy, nor chemistry [1] and the byline gives
    it away, /Money, Banking, and the Future of Global Economy/.

    Mervyn King was a member of the Royal Mint I believe, but it's been a
    while since I read it.

    Now, since you've read the book /The Black Swan/, how many black swan
    events have you caused?

    I've got three patents, which probably do qualify as black swan events
    - it the ideas had been predictable they would have been obvious to
    those skilled in the art.

    Congratulations-ish on that.


    Or do you think there's no correlation between
    reading a book, and causing an event?-a Can Karl Marx say the same [2]?

    Jonathon Israel thinks that Spinoza writing a couple of books had a
    lot to do with the democratic revolution.

    https://www.jstor.org/stable/j.ctt7t0f8

    I've added this book to my /saved for later/ on Abebooks.-a I won't do
    more for now.

    He's written more and longer books on that and related subjects. I got
    onto him when I moved to the Netherlands and read his history of the
    Dutch Republic.

    Books reflect the issues of the day, and can influence them. Karl Marx
    was a remarkably influential figure, and his ideas about "the leading
    role of the party" were unfortunately influential. Much of what he had
    to say was very sensible, but that lot wasn't.

    Anyway, don't stop on my account, you can still cause a few black swan
    events in your life, if you're not lazy.

    Laziness probably doesn't come into it. Black swan events are
    unpredictable, and if you can't predict what effect unexpected idea
    will have you can't put a lot of energy into encouraging the effects.

    [1] Which could be interesting enough, given your background in
    chemistry.

    It wasn't just my background. Both my parents had university degrees
    in chemistry,-a but nether was remotely interested in alchemy.

    Alchemy as the precursor to modern chemistry is quite fascinating.

    Only if you are fond of looking at intellectual errors.

    I did read up on Antoine Lavoisier, but only the superficial stuff
    readily available on the web.-a I haven't dug into any detailed > biographies yet.

    The Museum of Arts and Metiers in Paris has his chemistry lab on
    display. He was a tax farmer and made a lot of money out of it and his
    lab gear is remarkably gorgeous.

    Abu Musa Jabir ibn Hayyan is more famous for bringing us the term /gibberish/ with which I'm sure you're well acquainted.

    Strange you should bring that up. You do post a lot of it.

    [2] Well, technically, it's /writing a book/ for him, but that's just
    a minor difference, isn't it?

    The guy who ghost-wrote "The Art of the Deal" for Donald Trump had to
    put a lot of effort making what he wrote look as if Donald Trump might
    have written it, and he complained afterwards that it was difficult to
    get enough out of Donald Trump on the subject to create a document
    that might look as if Donald Trump had written it. A ninety-second
    attention span doesn't generate a lot of text in one hit, so the
    ghost-writer had to glom together a lot of sound-bites.


    Well, as I don't know Trump personally, I don't feel any need to bring
    him up in every conversation.-a I suggest you do the same.

    I'm not interested in /The Art of the Deal/, and not because Trump did
    not write it himself.-a It would be much more interesting /had/ he writ-
    ten it himself, I admit.-a I mean, the man must have learned to type at
    some point in life, if only to play /Leather Goddesses of Phobos/.

    I never finished the game myself, but I'm sure I'll get to it /one of
    these years/.-a Did you play it?

    I confine myself to Sid Meir's simulation games. My wife noticed that
    some of her more intelligent colleagues liked them and correctly deduced
    that I would.

    The lady who won an ignoble prize liked hers too much, and after playing
    it non-stop for 24 hours binned it and never touched it again.
    --
    Bill Sloman, Sydney

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Sloman@bill.sloman@ieee.org to sci.electronics.design on Sun Nov 15 15:47:57 2026
    From Newsgroup: sci.electronics.design

    On 15/11/2026 3:30 am, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/14/2026 10:14 PM, Bill Sloman wrote:
    On 14/11/2026 10:54 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/14/2026 3:20 PM, Bill Sloman wrote:
    On 14/11/2026 5:20 pm, Johann 'Myrkraverk' Oskarsson wrote:
    On 11/14/2026 10:04 AM, JM wrote:
    On Sat, 14 Nov 2026 07:44:45 +0800, Johann 'Myrkraverk' Oskarsson
    <johann@myrkraverk.invalid> wrote:

    <snip>

    You do seem to think that your own actions are consequential. JM and
    Phil Hobbs have been around here long enough to have seen any number
    of voluble half-wits run out of steam.

    You have nothing to worry about.-a I deleted my Steam account, so I won't >>> run out of Steam!-a I already did.

    The phrase is rather older than computer gaming, and refers to steam
    engines. Think George Stevensons's Rocket of 1829.

    https://en.wikipedia.org/wiki/Stephenson%27s_Rocket

    Yes, I know that.-a I just don't care.

    An easy claim to make after you have been shown up. Not actually convincing.
    --
    Bill Sloman, Sydney
    --- Synchronet 3.22a-Linux NewsLink 1.2