• Re: more Fortran, Comment delimiters

    From Bill Findlay@findlaybill@blueyonder.co.uk to comp.arch on Tue Oct 20 20:53:55 2026
    From Newsgroup: comp.arch

    On 20/10/2026 04:50, John Levine wrote:
    According to Thomas Koenig <tkoenig@netcologne.de>:
    DEC F77 used a stack and allowed recursion if you were careful
    whereas IBM F77 did not (it used static allocation for all local
    variables so recursion stomped on their values).

    OS/360 calling conventions is one thing they got wrong.

    My PDP-11 Unix F77 used the C compiler's back end, so recursion
    worked fine. I made variables automatic (stack) unless you used
    a SAVE statement, which made sense in the PDP-11's limited address space.

    It is quite possible to do recursion with the OS/360 calling convention. Their PL/I compilers did it and I believe the various ALGOL compilers,
    which had to support recursion, also used the standard calling sequence.


    Brian Wichmann (of Whetstone Benchmark fame) gives the following
    figures for Ackermann's function, a stress test for recursion.

    ALGOL-a60,-aby-aarchitecture,-amean orders/call,-a*-a=-aestimate

    RMCS, VAX 12.5
    XALGOL, B6700 16.0
    RMCS, PE 3200 17.5
    XALGOL B5500 19.5
    ICL, ICL 2900 Series 19.5
    Manchester Uni., ICL 1900 Series 33.5

    System/3?0, by language and compiler

    ALGOL 60, Edinburgh Uni. /360 21.0
    PL/I OPT v1.2.1, IBM /360 61 *
    ALGOL W, Stanford Uni. /360 74 *
    ALGOL 60, Delft Uni. /360 142 *
    PL/I F level v5.4, IBM /360 212 *
    ALGOL 60, IBM F level /360 820 *

    Make of that what you will.

    And for amusement:

    DEC VAX: Language, Compiler

    Ada, York Ada global function 8
    C optimised 9
    C 10
    Ada, York Ada nested function 11
    Ada, RMCS compiler 12.5

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Thomas Koenig@tkoenig@netcologne.de to comp.arch on Tue Oct 20 20:15:48 2026
    From Newsgroup: comp.arch

    Bill Findlay <findlaybill@blueyonder.co.uk> schrieb:
    On 20/10/2026 04:50, John Levine wrote:
    According to Thomas Koenig <tkoenig@netcologne.de>:
    DEC F77 used a stack and allowed recursion if you were careful
    whereas IBM F77 did not (it used static allocation for all local
    variables so recursion stomped on their values).

    OS/360 calling conventions is one thing they got wrong.

    My PDP-11 Unix F77 used the C compiler's back end, so recursion
    worked fine. I made variables automatic (stack) unless you used
    a SAVE statement, which made sense in the PDP-11's limited address space.

    It is quite possible to do recursion with the OS/360 calling convention.
    Their PL/I compilers did it and I believe the various ALGOL compilers,
    which had to support recursion, also used the standard calling sequence.


    Brian Wichmann (of Whetstone Benchmark fame) gives the following
    figures for Ackermann's function, a stress test for recursion.

    ALGOL-a60,-aby-aarchitecture,-amean orders/call,-a*-a=-aestimate

    I understand "ALGOL 60", but not the rest.

    What orders?
    --
    This USENET posting was made without artificial intelligence,
    artificial impertinence, artificial arrogance, artificial stupidity,
    artificial flavorings or artificial colorants.
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From scott@scott@slp53.sl.home (Scott Lurndal) to comp.arch on Tue Oct 20 20:44:45 2026
    From Newsgroup: comp.arch

    Thomas Koenig <tkoenig@netcologne.de> writes:
    Bill Findlay <findlaybill@blueyonder.co.uk> schrieb:
    On 20/10/2026 04:50, John Levine wrote:
    According to Thomas Koenig <tkoenig@netcologne.de>:
    DEC F77 used a stack and allowed recursion if you were careful
    whereas IBM F77 did not (it used static allocation for all local
    variables so recursion stomped on their values).

    OS/360 calling conventions is one thing they got wrong.

    My PDP-11 Unix F77 used the C compiler's back end, so recursion
    worked fine. I made variables automatic (stack) unless you used
    a SAVE statement, which made sense in the PDP-11's limited address space. >>>
    It is quite possible to do recursion with the OS/360 calling convention. >>> Their PL/I compilers did it and I believe the various ALGOL compilers,
    which had to support recursion, also used the standard calling sequence. >>>

    Brian Wichmann (of Whetstone Benchmark fame) gives the following
    figures for Ackermann's function, a stress test for recursion.

    ALGOL-a60,-aby-aarchitecture,-amean orders/call,-a*-a=-aestimate

    I understand "ALGOL 60", but not the rest.

    What orders?

    Archaic terminology for "instructions".

    The early Burroughs systems (Electrodata 220, Burroughs 300,
    B3500) all called them 'orders' rather than instructions.

    In fact, the processor test suite for the B3500 and successors
    was called all-orders.
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From quadibloc@quadibloc@invalid.com (John Savard) to comp.arch on Tue Oct 20 22:49:07 2026
    From Newsgroup: comp.arch

    On Tue, 20 Oct 2026 20:44:45 GMT, scott@slp53.sl.home (Scott Lurndal)
    wrote:

    Archaic terminology for "instructions".

    The early Burroughs systems (Electrodata 220, Burroughs 300,
    B3500) all called them 'orders' rather than instructions.

    In fact, the processor test suite for the B3500 and successors
    was called all-orders.

    At one time, the _instruction set_ of a computer was instead referred
    to as its "order code".

    John Savard
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From John Levine@johnl@taugh.com to comp.arch on Wed Oct 21 03:28:44 2026
    From Newsgroup: comp.arch

    According to Bill Findlay <findlaybill@blueyonder.co.uk>:
    System/3?0, by language and compiler

    ALGOL 60, Edinburgh Uni. /360 21.0
    PL/I OPT v1.2.1, IBM /360 61 *
    ALGOL W, Stanford Uni. /360 74 *
    ALGOL 60, Delft Uni. /360 142 *
    PL/I F level v5.4, IBM /360 212 *
    ALGOL 60, IBM F level /360 820 *

    That looks about right. PL/I F was the original compiler which generated lousy code, later replaced by PL/I OPT. ALGOL F seems to have been written with the attitude, you want an Algol compiler? Here's an Algol compiler, suck it up. It did GETMAIN and FREEMAIN system calls to allocate storage whenever it entered or
    left a block. Someone at Princeton patched it to do what PL/I did, allocate large blocks and suballocate within them, which vastly improved its performance.
    --
    Regards,
    John Levine, johnl@taugh.com, Primary Perpetrator of "The Internet for Dummies",
    Please consider the environment before reading this e-mail. https://jl.ly
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From quadibloc@quadibloc@invalid.com (John Savard) to comp.arch on Wed Oct 21 14:10:49 2026
    From Newsgroup: comp.arch

    On Wed, 21 Oct 2026 03:28:44 -0000 (UTC), John Levine
    <johnl@taugh.com> wrote:

    ALGOL F seems to have been written with the
    attitude, you want an Algol compiler? Here's an Algol compiler, suck it up.

    Yes. Unlike ALGOL W you had to indicate the keywords by putting quote
    marks around them - although some halfway-serious ALGOL 60 compilers
    for other architectures also did this.
    I presume that while ALGOL was not in any way a priority at IBM, they
    felt that some European customers, at least, might insist on an ALGOL
    compiler being available for the computer in order to consider
    purchasing a System/360.
    Given this, it's somewhat odd that the IBM 7030 or STRETCH went to the
    length of including the special characters of publication-level ALGOL
    in its character set. But perhaps in the years between the STRETCH and
    the 360, the bright hopes for ALGOL that may have existed at its
    inception had dimmed.

    John Savard
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From cross@cross@spitfire.i.gajendra.net (Dan Cross) to comp.arch on Wed Oct 21 20:11:30 2026
    From Newsgroup: comp.arch

    In article <6ad8c6dc.5749031@news.eternal-september.org>,
    John Savard <quadibloc@invalid.com> wrote:
    On Wed, 21 Oct 2026 03:28:44 -0000 (UTC), John Levine
    <johnl@taugh.com> wrote:

    ALGOL F seems to have been written with the
    attitude, you want an Algol compiler? Here's an Algol compiler, suck it up.

    Yes. Unlike ALGOL W you had to indicate the keywords by putting quote
    marks around them - although some halfway-serious ALGOL 60 compilers
    for other architectures also did this.

    I don't believe that is true. The descriptions of AGOL W that
    are publicly available do not indicate that quote marks were
    required around keywords; I have only seen that on Algol 60
    code.

    - Dan C.

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From cross@cross@spitfire.i.gajendra.net (Dan Cross) to comp.arch on Wed Oct 21 20:38:32 2026
    From Newsgroup: comp.arch

    In article <11bb69i$j4n$1@reader2.panix.com>,
    Dan Cross <cross@spitfire.i.gajendra.net> wrote:
    In article <6ad8c6dc.5749031@news.eternal-september.org>,
    John Savard <quadibloc@invalid.com> wrote:
    On Wed, 21 Oct 2026 03:28:44 -0000 (UTC), John Levine
    <johnl@taugh.com> wrote:

    ALGOL F seems to have been written with the
    attitude, you want an Algol compiler? Here's an Algol compiler, suck it up. >>
    Yes. Unlike ALGOL W you had to indicate the keywords by putting quote
    marks around them - although some halfway-serious ALGOL 60 compilers
    for other architectures also did this.

    I don't believe that is true. The descriptions of AGOL W that
    are publicly available do not indicate that quote marks were
    required around keywords; I have only seen that on Algol 60
    code.

    Oops, I misread: you said _unlike_ Algol W, and you are
    completely correct. Apologies!

    - Dan C.

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Findlay@findlaybill@blueyonder.co.uk to comp.arch on Wed Oct 21 23:06:04 2026
    From Newsgroup: comp.arch

    On 20/10/2026 23:49, John Savard wrote:
    On Tue, 20 Oct 2026 20:44:45 GMT, scott@slp53.sl.home (Scott Lurndal)
    wrote:

    Archaic terminology for "instructions".

    The early Burroughs systems (Electrodata 220, Burroughs 300,
    B3500) all called them 'orders' rather than instructions.

    In fact, the processor test suite for the B3500 and successors
    was called all-orders.

    At one time, the _instruction set_ of a computer was instead referred
    to as its "order code".
    The US-centrism is strong in these replies.
    Sigh. (I have taken that over from Nick.)
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From scott@scott@slp53.sl.home (Scott Lurndal) to comp.arch on Thu Oct 22 00:02:25 2026
    From Newsgroup: comp.arch

    Bill Findlay <findlaybill@blueyonder.co.uk> writes:
    On 20/10/2026 23:49, John Savard wrote:
    On Tue, 20 Oct 2026 20:44:45 GMT, scott@slp53.sl.home (Scott Lurndal)
    wrote:

    Archaic terminology for "instructions".

    The early Burroughs systems (Electrodata 220, Burroughs 300,
    B3500) all called them 'orders' rather than instructions.

    In fact, the processor test suite for the B3500 and successors
    was called all-orders.

    At one time, the _instruction set_ of a computer was instead referred
    to as its "order code".
    The US-centrism is strong in these replies.
    Sigh. (I have taken that over from Nick.)

    Perhaps you might, then, enlighten your yank
    cousins by providing the UK viewpoint.
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Findlay@findlaybill@blueyonder.co.uk to comp.arch on Thu Oct 22 23:43:15 2026
    From Newsgroup: comp.arch

    On 22/10/2026 01:02, Scott Lurndal wrote:
    Bill Findlay <findlaybill@blueyonder.co.uk> writes:
    On 20/10/2026 23:49, John Savard wrote:
    On Tue, 20 Oct 2026 20:44:45 GMT, scott@slp53.sl.home (Scott Lurndal)
    wrote:

    Archaic terminology for "instructions".

    The early Burroughs systems (Electrodata 220, Burroughs 300,
    B3500) all called them 'orders' rather than instructions.

    In fact, the processor test suite for the B3500 and successors
    was called all-orders.

    At one time, the _instruction set_ of a computer was instead referred
    to as its "order code".
    The US-centrism is strong in these replies.
    Sigh. (I have taken that over from Nick.)

    Perhaps you might, then, enlighten your yank
    cousins by providing the UK viewpoint.

    I do, from time to time, and am generally ignored.
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From MitchAlsup@user5857@newsgrouper.org.invalid to comp.arch on Fri Oct 23 00:27:07 2026
    From Newsgroup: comp.arch

    Bill Findlay <findlaybill@blueyonder.co.uk> posted:

    On 22/10/2026 01:02, Scott Lurndal wrote:
    Bill Findlay <findlaybill@blueyonder.co.uk> writes:
    On 20/10/2026 23:49, John Savard wrote:
    On Tue, 20 Oct 2026 20:44:45 GMT, scott@slp53.sl.home (Scott Lurndal)
    wrote:

    Archaic terminology for "instructions".

    The early Burroughs systems (Electrodata 220, Burroughs 300,
    B3500) all called them 'orders' rather than instructions.

    In fact, the processor test suite for the B3500 and successors
    was called all-orders.

    At one time, the _instruction set_ of a computer was instead referred
    to as its "order code".
    The US-centrism is strong in these replies.
    Sigh. (I have taken that over from Nick.)

    Perhaps you might, then, enlighten your yank
    cousins by providing the UK viewpoint.

    I do, from time to time, and am generally ignored.

    That is ALMOST the Casandra Syndrome:: "always tell the truth,
    nobody ever believes"
    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Bill Findlay@findlaybill@blueyonder.co.uk to comp.arch on Fri Oct 23 03:46:14 2026
    From Newsgroup: comp.arch

    On 23/10/2026 01:27, MitchAlsup wrote:
    Bill Findlay <findlaybill@blueyonder.co.uk> posted:

    On 22/10/2026 01:02, Scott Lurndal wrote:
    Bill Findlay <findlaybill@blueyonder.co.uk> writes:
    On 20/10/2026 23:49, John Savard wrote:
    On Tue, 20 Oct 2026 20:44:45 GMT, scott@slp53.sl.home (Scott Lurndal) >>>>> wrote:

    Archaic terminology for "instructions".

    The early Burroughs systems (Electrodata 220, Burroughs 300,
    B3500) all called them 'orders' rather than instructions.

    In fact, the processor test suite for the B3500 and successors
    was called all-orders.

    At one time, the _instruction set_ of a computer was instead referred >>>>> to as its "order code".
    The US-centrism is strong in these replies.
    Sigh. (I have taken that over from Nick.)

    Perhaps you might, then, enlighten your yank
    cousins by providing the UK viewpoint.

    I do, from time to time, and am generally ignored.

    That is ALMOST the Casandra Syndrome:: "always tell the truth,
    nobody ever believes"

    Quite so.
    However, it all goes into an archive and will continue
    being available to those with curiosity and broad minds.

    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From Michael S@already5chosen@yahoo.com to comp.arch on Fri Oct 23 11:21:15 2026
    From Newsgroup: comp.arch

    On Fri, 23 Oct 2026 03:46:14 +0100
    Bill Findlay <findlaybill@blueyonder.co.uk> wrote:

    On 23/10/2026 01:27, MitchAlsup wrote:
    Bill Findlay <findlaybill@blueyonder.co.uk> posted:

    On 22/10/2026 01:02, Scott Lurndal wrote:
    Bill Findlay <findlaybill@blueyonder.co.uk> writes:
    On 20/10/2026 23:49, John Savard wrote:
    On Tue, 20 Oct 2026 20:44:45 GMT, scott@slp53.sl.home (Scott
    Lurndal) wrote:

    Archaic terminology for "instructions".

    The early Burroughs systems (Electrodata 220, Burroughs 300,
    B3500) all called them 'orders' rather than instructions.

    In fact, the processor test suite for the B3500 and successors
    was called all-orders.

    At one time, the _instruction set_ of a computer was instead
    referred to as its "order code".
    The US-centrism is strong in these replies.
    Sigh. (I have taken that over from Nick.)

    Perhaps you might, then, enlighten your yank
    cousins by providing the UK viewpoint.

    I do, from time to time, and am generally ignored.

    That is ALMOST the Casandra Syndrome:: "always tell the truth,
    nobody ever believes"

    Quite so.
    However, it all goes into an archive and will continue
    being available to those with curiosity and broad minds.


    Archives?
    Google's usenet archives became very unpleasant to use in recent months.
    Are there alternatives?




    --- Synchronet 3.22a-Linux NewsLink 1.2
  • From David Brown@david.brown@hesbynett.no to comp.arch on Fri Oct 23 12:49:05 2026
    From Newsgroup: comp.arch

    On 23/10/2026 04:46, Bill Findlay wrote:
    On 23/10/2026 01:27, MitchAlsup wrote:
    Bill Findlay <findlaybill@blueyonder.co.uk> posted:

    On 22/10/2026 01:02, Scott Lurndal wrote:
    Bill Findlay <findlaybill@blueyonder.co.uk> writes:
    On 20/10/2026 23:49, John Savard wrote:
    On Tue, 20 Oct 2026 20:44:45 GMT, scott@slp53.sl.home (Scott Lurndal) >>>>>> wrote:

    Archaic terminology for "instructions".

    The early Burroughs systems (Electrodata 220, Burroughs 300,
    B3500) all called them 'orders' rather than instructions.

    In fact, the processor test suite for the B3500 and successors
    was called all-orders.

    At one time, the _instruction set_ of a computer was instead referred >>>>>> to as its "order code".
    The US-centrism is strong in these replies.
    Sigh. (I have taken that over from Nick.)

    Perhaps you might, then, enlighten your yank
    cousins by providing the UK viewpoint.

    I do, from time to time, and am generally ignored.

    That is ALMOST the Casandra Syndrome:: "always tell the truth,
    nobody ever believes"

    Quite so.
    However, it all goes into an archive and will continue
    being available to those with curiosity and broad minds.


    What do you think is going in the archives here? All I see is an unsubstantiated claim of "US centrism" for something that doesn't really matter very much anyway. I have no idea what terms were used at that
    time, so I have no idea if you are right or not - but you should justify
    your accusations with verifiable facts, or at the very least, anecdotes
    and personal experience.

    I am quite confident that our friends across the pond are not being intentionally US-centric about their history, but it is natural that
    their own experiences and what they have heard is somewhat biased - just
    like it is for the rest of us. I occasionally get corrected for my UK
    bias of history, and I appreciate such corrections - I think most people appreciate corrections more than unhelpful accusations.

    --- Synchronet 3.22a-Linux NewsLink 1.2