| Sysop: | Amessyroom |
|---|---|
| Location: | Fayetteville, NC |
| Users: | 27 |
| Nodes: | 6 (0 / 6) |
| Uptime: | 40:28:01 |
| Calls: | 631 |
| Calls today: | 2 |
| Files: | 1,187 |
| D/L today: |
24 files (29,813K bytes) |
| Messages: | 174,392 |
Den 29.09.2025 kl. 23.48 skrev Silvano:
You forgot a smiley and you can say anything you want, but you can't
force me to listen to what you're saying. That's why I have scores of
killfiles, over 100 in a German group.
I think that you have entries that are no longer needed. I have a total
of four identities that I kill in desd.
On 30/09/2025 10:51, Athel Cornish-Bowden wrote:
On 2025-09-29 21:12:29 +0000, Garrett Wollman said:
In article <10begad$35mtr$1@dont-email.me>,
Athel Cornish-Bowden <me@yahoo.com> wrote:
Increasingly often (and it seems to have increased
dramatically in the
last week or so) I get a dialogue demanding that I show that I
am not a
robot, with the message "Verifying you are human. This may
take a few
seconds." or something similar.
But it just goes away in a second, without you having to do
anything,
right?
That's an anti-"AI"-scraper countermeasure. With billions of
dollars
being spent on training "AI" models, the companies that make
them are
desparate for any training corpus they can get, and have been
scraping
hitting web site on earth at such speed and volume that sites
that do
not otherwise require login are melting (or racking up very large
provider bills) under the pressure. So the site owners are
fighting
back by doing a bit of profiling to simultaneously slow down the
scrapers, force them to expend commensurate computational
effort to
the sites they're scraping, and potentially profile them to
serve them
different content (like a "not authorized" error). It used to
be that
there were better mechanisms to do this, like IP address
reputation
systems, but some of the "AI" companies are effectively renting
botnets to make it more difficult to block their scraping,
which makes
it nearly impossible for anything other than a behavior-based
system
to effectively exclude the parasites.
Thanks. That makes a lot of sense, and explains why this has
happened so suddenly.
Why is it that, whenever we find a really good new way to put
computers to make life better for everyone, we quickly find a way
to use it to make life worse instead?
On 2025-09-29 21:12:29 +0000, Garrett Wollman said:
Athel Cornish-Bowden-awrote:
Increasingly often (and it seems to have increased dramatically in the
last week or so) I get a dialogue demanding that I show that I am not a
robot, with the message "Verifying you are human. This may take a few
seconds." or something similar.
But it just goes away in a second, without you having to do anything,
right?
That's an anti-"AI"-scraper countermeasure.-a With billions of dollars
being spent on training "AI" models, the companies that make them are
desparate for any training corpus they can get, and have been scraping
hitting web site on earth at such speed and volume that sites that do
not otherwise require login are melting (or racking up very large
provider bills) under the pressure.-a So the site owners are fighting
back by doing a bit of profiling to simultaneously slow down the
scrapers, force them to expend commensurate computational effort to
the sites they're scraping, and potentially profile them to serve them
different content (like a "not authorized" error).-a It used to be that
there were better mechanisms to do this, like IP address reputation
systems, but some of the "AI" companies are effectively renting
botnets to make it more difficult to block their scraping, which makes
it nearly impossible for anything other than a behavior-based system
to effectively exclude the parasites.
Thanks. That makes a lot of sense, and explains why this has happened so suddenly.
I forgot to mention that killfiles are, of course, a violation of free speech.
On Mon, 29 Sep 2025 22:53:10 +0100, Aidan Kehoe <kehoea@parhasard.net>
wrote:
consciously make the effort to avoid having it attempt to mislead me, e.g. by >>interpolating ideas from other jurisdictions into Irish tax law, what I am >>interested in. Search for a PDF manual of a device you have to debug (e.g. my >>surgeryrCOs phone system); this used to work twenty years ago, now you get page
after page after page of boilerplate content that soaks up clicks and attention
without actually supplying the manual.
That seems odd to me. I've downloaded a half dozen manuals (for
tv, phone, stereo) and don't remember trouble in finding them.
At worst, the manufacturer's site wants me to specify the type
of appliance and supply the model name/ number.
Den 30.09.2025 kl. 02.48 skrev lar3ryca:
I forgot to mention that killfiles are, of course, a violation of
free speech.
No it isn't. Free speech means that I cannot prevent someone from
speaking. It does NOT mean that I have to listen to them.
Not even if they repeat themselves.
On 30/09/2025 03:14, lar3ryca wrote:
<snip>
I am SO happy that I seldom see those stupid picture tests any more.
Same for the ones that ask you to type in the distorted characters.
The early distorted characters were not all made-up distortions. Luis
von Ahn, the inventor of the CAPTCHA system (https://en.wikipedia.org/wiki/Luis_von_Ahn) explains that they [the
CAPTCHA inputs] were used to 'debug' the scans of old documents (e.g
Google Book project), where the OCR scanner had failed. So, in order to resolve the problem, his system used human skills - disguised as
CAPTCHA texts - to correct the OCR results. (There was quite a lot of correlation of typed responses to come to a consensus verdict.)
P.S: Luis von Ahn is the co-founder and CEO of Duolingo.
What I mostly see now are the ones that show a box and ask if you are
human. By the time I get to the box, it has usually decided I am human.
I think, but can't be sure, that they are testing reaction times, or
perhaps mouse movements. If I can remember, I'll try not moving my mouse
at all to see if that changes things.
I've googled, but it's hard to find anything concrete. I expect the algorithms are kept secret to try to curb the arms race between robots
and no-bots. Google's AI (for what it's worth) says: "Cloudflare
verifies humans by analyzing user behavior, browser information, and
device characteristics to distinguish real users from bots. This
process often involves a 'JavaScript challenge' where the browser must successfully execute complex JavaScript code, or it may use a CAPTCHA
with image puzzles. For a more seamless, background process,
Cloudflare's Turnstile technology uses cryptographic attestation to
verify a device's trusted status without requiring user interaction or visible challenges. ... Cryptographic Attestation ... verifies that
the user's device is controlled by a legitimate entity by checking for
a public key signed by a trusted manufacturer...."
Rich Ulrich <rich.ulrich@comcast.net> wrote or quoted:
I don't regret creating a second Google icon, one that gives me the
search without AI. I ignored it for a while, but I've been using it
more in recent weeks.
If you know how to write a bit of JavaScript, you can do a
ton with Google results pages. For example, you can strip out
everything except the actual search hits. Then you just stash
that as a bookmarklet and fire it up with a couple key presses.
Den 29.09.2025 kl. 23.48 skrev Silvano:
You forgot a smiley and you can say anything you want, but you
can't force me to listen to what you're saying. That's why I have
scores of killfiles, over 100 in a German group.
I think that you have entries that are no longer needed. I have a
total of four identities that I kill in desd.
I still miss the feature of Agent. You could set a kill filter to
be active X days after the last hit.
Den 30.09.2025 kl. 07.16 skrev Steve Hayes:
My older version of Firefox had its own cookie popup, which was
much better. I could decide to allow cookies always, or allow
them to the end of the session. If it was a site I visited
regularly, I'd click on "Allow" and if it was a once-off visit,
I'd toke the "End of session" option.
Unfortunately newer versions don't seem to have that option.
In the preferences I have set up that cookies will disappear when
I close the browser, so I always accept the very few times when "I
don't care about cookies" hasn't removed the disturbance. I can
also enter exceptions for sites whose cookies will survive. I
don't miss the choice options.
On 30 Sep 2025, Bertel Lund Hansen wrote
In the preferences I have set up that cookies will disappear when I
close the browser, so I always accept the very few times when "I
don't care about cookies" hasn't removed the disturbance. I can
also enter exceptions for sites whose cookies will survive. I don't
miss the choice options.
Same here, except that I don't enter exceptions for any sites -- I
just wipe all cookies, trackers, and such-like (along with clearing
the browser history, which I've simply never, ever used in the 30
years or so that I've had internet access). I keep saved passwords
for logins, though.
I still miss the feature of Agent. You could set a kill filter toThat feature still exists in XNews (which also, amazingly, still
be active X days after the last hit.
exists). It offers a default expiry of 100 days, which you can change
to anything you like ("0"="never").
In the preferences I have set up that cookies will disappear when
I close the browser, so I always accept the very few times when "I
don't care about cookies" hasn't removed the disturbance. I can
also enter exceptions for sites whose cookies will survive. I
don't miss the choice options.
Same here, except that I don't enter exceptions for any sites -- I
just wipe all cookies, trackers, and such-like (along with clearing
the browser history, which I've simply never, ever used in the 30
years or so that I've had internet access). I keep saved passwords
for logins, though.
It means I have to approach all sites as if I'm a new visitor each
time I go there (which I am, as far as they're concerned), but I'm
used to that; it doesn't bother me.
In hindsight, cookies should never have been invented. They have
relatively few legitimate uses, and a whole lot of evil uses.
Den 01.10.2025 kl. 13.38 skrev Peter Moylan:
In hindsight, cookies should never have been invented. They have
relatively few legitimate uses, and a whole lot of evil uses.
Isn't it true that sessions in PHP (and similar web programming
languages) only work because they set cookies?
In hindsight, cookies should never have been invented. They have
relatively few legitimate uses, and a whole lot of evil uses.
Isn't it true that sessions in PHP (and similar web programming
languages) only work because they set cookies?
Yes.
Den 01.10.2025 kl. 13.27 skrev HVS:
In the preferences I have set up that cookies will disappear when
I close the browser, so I always accept the very few times when "I
don't care about cookies" hasn't removed the disturbance. I can
also enter exceptions for sites whose cookies will survive. I
don't miss the choice options.
Same here, except that I don't enter exceptions for any sites -- I
just wipe all cookies, trackers, and such-like (along with clearing
the browser history, which I've simply never, ever used in the 30
years or so that I've had internet access).-a I keep saved passwords
for logins, though.
I also clear my history when I close the browser. The cookie exceptions
I have set concern the newspaper that I read and a site with our authoritative dictionary of spelling. I've forgotten why I allowed that
last one, but there was a reason.
It means I have to approach all sites as if I'm a new visitor each
time I go there (which I am, as far as they're concerned), but I'm
used to that; it doesn't bother me.
I have only found one example of a functionality that I can't have:
There's a site with a tv-program for many channels. One can set up that
only one's favourite channels are shown, but I can't[1]. I can live with that.
[1] Well, I can, but the setup is gone when I revisit.
Den 01.10.2025 kl. 18.24 skrev Richard Heathfield:
In hindsight, cookies should never have been invented. They have
relatively few legitimate uses, and a whole lot of evil uses.
Isn't it true that sessions in PHP (and similar web
programming languages) only work because they set cookies?
Yes.
That means that a whole lot of user-friendly webpages would be
destroyed if we removed cookies.
Yes, but Peter wasn't suggesting that; he was just bemoaning
their invention. If they had never been invented, PHP & co would
simply have found a way to store session information server side.
In article <10bjujt$icsr$1@dont-email.me>,
Richard Heathfield <rjh@cpax.org.uk> wrote:
Yes, but Peter wasn't suggesting that; he was just bemoaning
their invention. If they had never been invented, PHP & co would
simply have found a way to store session information server side.
That information can't *all* be stored on the server side - the server
has to at least have some way to differentiate clients.
On 01/10/2025 19:19, Bertel Lund Hansen wrote:
Den 01.10.2025 kl. 18.24 skrev Richard Heathfield:
In hindsight, cookies should never have been invented. They have
relatively few legitimate uses, and a whole lot of evil uses.
Isn't it true that sessions in PHP (and similar web programming
languages) only work because they set cookies?
Yes.
That means that a whole lot of user-friendly webpages would be destroyed if we removed cookies.
Yes, but Peter wasn't suggesting that; he was just bemoaning their invention. If they had never been invented, PHP & co would simply have found a way to store session information server side.
That information can't *all* be stored on the server side - the server
has to at least have some way to differentiate clients.
Login.
In article <10bk1tm$icsq$1@dont-email.me>,
Richard Heathfield <rjh@cpax.org.uk> wrote:
That information can't *all* be stored on the server side - the server
has to at least have some way to differentiate clients.
Login.
What exactly do you mean by that? Typical remote computer logins last
for the duration of a TCP connection, while HTTP is designed not to
rely on persistent connections. It could be done,
but it's not an
alternative to cookies if you're using the WWW.
In article <10bk1tm$icsq$1@dont-email.me>,
Richard Heathfield <rjh@cpax.org.uk> wrote:
That information can't *all* be stored on the server side - the server
has to at least have some way to differentiate clients.
Login.
What exactly do you mean by that? Typical remote computer logins last
for the duration of a TCP connection, while HTTP is designed not to
rely on persistent connections. It could be done, but it's not an alternative to cookies if you're using the WWW.
Isn't it true that sessions in PHP (and similar web programming
languages) only work because they set cookies?
Yes.
That means that a whole lot of user-friendly webpages would be
destroyed if we removed cookies.
Yes, but Peter wasn't suggesting that; he was just bemoaning their invention. If they had never been invented, PHP & co would simply have
found a way to store session information server side.
Den 01.10.2025 kl. 21.15 skrev Richard Heathfield:
Isn't it true that sessions in PHP (and similar web
programming languages) only work because they set cookies?
Yes.
That means that a whole lot of user-friendly webpages would be
destroyed if we removed cookies.
Yes, but Peter wasn't suggesting that; he was just bemoaning
their invention. If they had never been invented, PHP & co
would simply have found a way to store session information
server side.
But couldn't serverside cookies store the same information as
normal cookies? If so - where is the advantage?
richard@cogsci.ed.ac.uk (Richard Tobin) writes:
In article <10bk1tm$icsq$1@dont-email.me>, Richard Heathfield
<rjh@cpax.org.uk> wrote:
That information can't *all* be stored on the server side - the
server has to at least have some way to differentiate clients.
Login.
What exactly do you mean by that? Typical remote computer logins
last for the duration of a TCP connection, while HTTP is designed
not to rely on persistent connections. It could be done, but it's
not an alternative to cookies if you're using the WWW.
AIUI, the big selling point of HTTP was that it allowed document
retrieval without the overhead for the server of maintining the
login state. Connect, optional inline PW validation, fetch,
disconnect.
But soon (Immediately? When, in the time-line?) the demand for a
maintained session arose. Cookies.
But couldn't serverside cookies store the same information as normal >cookies? If so - where is the advantage?
TLS, which is what HTTPS uses, does have an option to restore a previous >session, but as far as I know nobody has yet used that for anything
except restoring the encryption state.
Perhaps we should develop a mechanism by which the owner of a web server
is required to pay for being allowed to store a cookie.
Increasingly often (and it seems to have increased dramatically in the last week or so) I get a dialogue demanding that I show that I am not a robot, with the message "Verifying you are human. This may take a few seconds." or something similar. Apart from being very irritating, it seems completely pointless: are they really so moronic that it doesn't occur to them that people who design robots are unable to get them to recognize such dialogues and to click in the little square?
Just as irritating, maybe more so, are sites that say that they've detected that I'm using an ad-blocker, and please disable it. I don't disable my ad-blocker but just look for another site. Amazing as it may seem, the reason
I use an ad-blocker is that I don't want to clutter up my screen with advertisements.
Soon we'll have trolls posting messages that say "I see that I'm in your killfile. Please let my message through as it's necessary for my trollish purposes."
no introduction needed:
<URL:https://www.kevinandkell.com/2016/kk0313.html>
(Safe for home use)
On 07/10/2025 02:27, Snidely wrote:
no introduction needed:
<URL:https://www.kevinandkell.com/2016/kk0313.html>
(Safe for home use)
<smile>
On 07/10/2025 07:02, occam wrote:
On 07/10/2025 02:27, Snidely wrote:
no introduction needed:
<URL:https://www.kevinandkell.com/2016/kk0313.html>
(Safe for home use)
<smile>
I assume you were not smiling at the humungous long list of cookies you
were 'invited' to accept before being given access to the cartoon?
Sam Plusnet hat am 07.10.2025 um 20:37 geschrieben:
On 07/10/2025 07:02, occam wrote:
On 07/10/2025 02:27, Snidely wrote:
no introduction needed:
<URL:https://www.kevinandkell.com/2016/kk0313.html>
(Safe for home use)
<smile>
I assume you were not smiling at the humungous long list of cookies you
were 'invited' to accept before being given access to the cartoon?
Just curious. WHAT? I didn't see any list! Otherwise I'd have gone away.
I wonder. Did I accept all cookies unwittingly? I don't think so.
Usually I'm not that stupid. Or does the fact that my computer is in the
EU make a difference?
On Tue, 7 Oct 2025 22:54:16 +0200, Silvano
<Silvano@noncisonopernessuno.it> wrote:
Sam Plusnet hat am 07.10.2025 um 20:37 geschrieben:
On 07/10/2025 07:02, occam wrote:
On 07/10/2025 02:27, Snidely wrote:
no introduction needed:
<URL:https://www.kevinandkell.com/2016/kk0313.html>
(Safe for home use)
<smile>
I assume you were not smiling at the humungous long list of cookies you
were 'invited' to accept before being given access to the cartoon?
Just curious. WHAT? I didn't see any list! Otherwise I'd have gone away.
I wonder. Did I accept all cookies unwittingly? I don't think so.
Usually I'm not that stupid. Or does the fact that my computer is in the
EU make a difference?
When a link is opened, a notice often pops up making you choose to
accept cookies or some cookies to view the site. The link produces no
such pop-up. There was no invitation that I saw.
On Tue, 7 Oct 2025 22:54:16 +0200, Silvano
<Silvano@noncisonopernessuno.it> wrote:
Sam Plusnet hat am 07.10.2025 um 20:37 geschrieben:
On 07/10/2025 07:02, occam wrote:
On 07/10/2025 02:27, Snidely wrote:
no introduction needed:
<URL:https://www.kevinandkell.com/2016/kk0313.html>
(Safe for home use)
<smile>
I assume you were not smiling at the humungous long list of cookies you
were 'invited' to accept before being given access to the cartoon?
Just curious. WHAT? I didn't see any list! Otherwise I'd have gone away.
I wonder. Did I accept all cookies unwittingly? I don't think so.
Usually I'm not that stupid. Or does the fact that my computer is in the
EU make a difference?
When a link is opened, a notice often pops up making you choose to
accept cookies or some cookies to view the site. The link produces no
such pop-up. There was no invitation that I saw.
Sam Plusnet hat am 07.10.2025 um 20:37 geschrieben:
On 07/10/2025 07:02, occam wrote:
On 07/10/2025 02:27, Snidely wrote:
no introduction needed:
<URL:https://www.kevinandkell.com/2016/kk0313.html>
(Safe for home use)
<smile>
I assume you were not smiling at the humungous long list of cookies you
were 'invited' to accept before being given access to the cartoon?
Just curious. WHAT? I didn't see any list! Otherwise I'd have gone away.
I wonder. Did I accept all cookies unwittingly? I don't think so.--
Usually I'm not that stupid. Or does the fact that my computer is in the
EU make a difference?