We have seen a single obnoxious poster who posts long Italian posts ranting about murderers and other criminals. The posts are easily recognised because they are all in capitals with very long titles.
As a group we have reported these postings as abuse and then they get hidden from the ordinary overview. For this sanitary action we seem to have been punished.
This same spam has been going for years, but the poster keeps changing their identity. It's always all-caps, lots of references to mafia, assasins, paedofilia etc.
> But I doubt google ever cared about Usenet at all
Doubt? This is without doubt: For-profit companies cares about generating profits. If they own something that doesn't generate profits, the only reason they even own it is because they haven't figured out how to close it down yet, or how to earn profits on it, or they think it leads to profits in other areas.
Of course they don't care about making information more widely available for free, they care about making money. Hardly surprising that they prefer to spend the least amount of time on moderating, closing the entire group is way simpler and cost-effective (according to them, apparently).
Wow if getting a group banned is that easy, then it's more like a vulnerability because the same can be scaled up and get almost any group banned (as the banning is also automatic as I see in another comment below).
Oh God, here we go again! I confirm that by searching on lumendatabase[0] requests to remove defamation posts are found;
again these requests are very specific and in Italian, so I suspect it is from the same victim of the stalker/harasser, so I suppose that ONCE AGAIN[1] instead of removing individual defamatory posts, google has 'decided' to nuke the entire group.
For the sake of completeness, the last time I couldn't get anything out of, the whole experience with multinationals of a certain calibre nowaday is to talk to cheap and dummy bots or worse: people who are so underpaid or unprepared or uninterested in solving the problem that they don't even try to understand it.
Well, that's a big one. Apparently the stalker annoyed the wrong person.
This[2] "Court Order Complaint to Google" was compiled and sent by Mediolanum (yes that mediolanum, Berlusconi's company shell), I have downloaded the attached documents and they actually are court orders and biglawyers' stuff... I don't think it's over.
So one mentally unstable individual has enough time and energy to make defamatory accusations (pedophilia, etc.) against this lawyer on multiple forums and in response the lawyer is sending these defamation complaints resulting in the nuking.
What is the word for painting a target on an innocent bystander so it gets bombed by the opposition?
I downloaded the court order, it's dated Sept. 2013, and the only reference to comp.lang.tcl is one entry in a list of 51 urls - https://fcku.it/comp.lang.tcl/thread/4177846 - not even accessing it via google groups!
Google bought dejanews, the historical usenet archive. They made it accessible via google groups. Now braindamaged bots are closing these archives. If google cannot fullfill their dejanews buyout promises they should hand it over to entities which are capable hosting these archives.
Not a single major newsgreoup like comp.lang.* may be closed by bots!
Year after year after year this same italian spammer on all comp-groups. What is the issue here?
Subject: RAPISCE, INCULA E AMMAZZA I BAMBINI: ANDREA BAGGIO DI REPUTATIONUP CASTELFRANCO VENETO (TV)!
From: Andreas Nigg <andreasnigg.safrasarasin@gmail.com>
What exactly do you expect as answer to this question? That a native italian speaker comes up with a peer reviewed proof that the text is, in fact, written in their mother tongue?
Italian is fairly easy to detect. Even though I don't know Italian, that's clearly Italian.
You can also test this by putting Google Translate / DeepL Translate into "automatic mode" and putting "RAPISCE, INCULA E AMMAZZA I BAMBINI: ANDREA BAGGIO DI REPUTATIONUP CASTELFRANCO VENETO" to translate, both of them correctly identifies it as Italian.
Unless there's any good reason to suspect otherwise, why not go with the "this is an italian poster, which is why they are writing in italian", as opposed to a multi-ligual person using a non-native language in a false flag operation.
Especially since genuine loonies posting such BS is quite common in forums, comment sections, and newsgroups and we all have seen examples.
Yeah, what motivation would there be to pretend to be Italian?
It's not like one crazy person is going to discredit Italy or Italians, and the poster is not going to be any more likely to be succesful at spamming because it's in Italian, or reach a wider audience.
'the first american, columbus' the fuck are you talking about? so all the people that were already living on the continent just don't count? Fuck Columbus.
Only because in the current culture “Italian” isn’t generally considered to be a “race”.
But there’s nothing fundamentally distinguishing “of Italian descent” from the types of groups which are considered “races”.
If (hypothetically) someone were to hate people on the basis of those people being of Italian descent, this would be quite similar to if they were racist in a specific sense towards some group which is considered a “race”.
However, presently, it would, I think, be significantly less harmful at the margin, because there are much fewer people that feel that way, and the overall harm scales super-linearly with the number-or-proportion-or-something of people who have the given prejudice.
That being said, I’m pretty sure that guy is trolling, and his complaint of anti-Italian racism is presumably insincere.
Google used to be the benefactor of the old internet and open source. Then, as soon as we got hooked on those free services, Gizmo turns into a Gremlin that smashes all our old toys so there can be no turning back.
> Google used to be the benefactor of the old internet and open source.
Was it? I can remember when it was a good search engine which let me find these, but I don’t think I ever thought “thank god google took this service over”, even back when they were building free services worth using 15 years ago.
Well, I guess it's up for interpretation, but for me "old internet" was when people ran their own email servers and their own domains. With that perspective, Gmail basically ruined being able to run your own email service as it's hard to get through the spam filter. Maybe it's better now but it used to be hard.
And not sure what Gmail contributed to open source, I can think of 0 projects/libraries/protocols/specifications that comes from Gmail and is used elsewhere widely. I guess that you can export to mbox via Google Takeout is something, but has very little to do with open source.
For me, 'old internet' is when individuals hand-crafted sites on the internet for their hobbies and interests, and searching Google actually surfaced those sites. Now, no one makes a website except to try to make money somehow.
So do I: A still active poster here on HN. (I think he was a Google employee at the time, which was why he had invites to dole out.) Around... 2004? Thereabouts.
Google's mission statement is to “organize the world's information and make it universally accessible and useful.” Its vision statement is to “provide an important service to the world-instantly delivering relevant information on virtually any topic.”
Google Groups needs to die. Or at least permanently dissociate from Usenet.
Every time they pull this temporary ban, the quality of the affected newsgroups improves.
The most recent stunt pulled by GG is that some genius decided that punctuation characters should be dropped from newsgroup names when posting. So now when GG users post to comp.lang.c++, the post goes to comp.lang.c.
GG does not provide a useablee Usenet archive. GG's net contribution to the Usenet ecosystem is negative.
They should just release the whole archive of Usenet posts that they acquired way back when into the public domain, and shutter the whole thing. They would do a favour to both themselves and digital preservationists.
And I remember searching and finding an answer from myself. Then I realised all data we generate will be sucked up and remembered for as long as it is considered valueable.
Me too and back then I'd never have imagined that it could essentially disappear. It's a bit as imaging stackoverflow to disappear. People will say there are the data dumps, someone will surely set up a mirror, etc. but in the end easily searchable and actively maintained content is something entirely different.
At least the 'easily searchable' part of the old Usenet content is something that might come back with https://www.usenetarchives.com/
Re the last link, as long as an archive only exists in a single place or is only operated by a single entity, it may as well not exist at all, because eventually it won’t. (This is also what terrifies me about the Wayback Machine and, to a much lesser extent, WinWorld.) An archivist who does not provide for easy mirroring is thus irresponsible.
I did something similar. It was that day I tried to make sure I use some alias when posting. What is benign one day will be heretical tomorrow and vice versa.
The "comp.windows.news" newsgroup about the NeWS window system was cursed with spam and offtopic posts, because so many people confused it with "news" about Microsoft Windows. Maybe if Usenet group names weren't required to be lower case, it could have been comp.windows.NeWS, and there would have been less confusion and spam.
This call for deletion describes its history of confusion and moderation:
The last approved article was in March 2000. Sun Microsystems
announced the end of life for the Network extensible Window System
(NeWS) in the early 1990's.
During the early 1990's, traffic in the then unmoderated
comp.windows.news was fairly light, less than 10 articles per week. In
December 1995, the group was moderated in place in order to exclude
misplaced articles related to "news" about MS-Windows. After the
moderation, fewer than 20 articles were approved to the newsgroup.
Because of its name, conversion to an unmoderated status is not
practical. An unmoderated newsgrooup would be overrun by MS-Windows
traffic (see comp.windows.misc).
HISTORY:
The unmoderated newsgroup comp.windows.news was created in April 1987,
soon after the Great Renaming. comp.windows.misc and comp.windows.x
are of similar vintage.
The newsgroup comp.windows.ms was created in November 1988 for
discussion of early versions of MS-Windows. Because of continued off
topic posting to the other newsgroups in the comp.windows.* hierarchy,
comp.windows.ms was re-organized into comp.os.ms-windows.* in 1992.
Despite creation of an entire hierarchy dedicated to MS-Windows,
off-topic posting continued in the comp.windows.* hierarchy. The
newsgroup comp.windows.news had the additional burden of a name that
suggested it was for "news" about (MS)-Windows, rather than being
about the NeWS.
In December 1995, comp.windows.news was converted to a moderated group
in order to exclude off-topic discussion, as well as "debate" about
the group name.
Over the next 5 years less than 20 articles were approved. The last
approval was in March 2000.
Am I the only one finding the popular usage of the word "Algorithm" annoyingly misleading? It's almost certainly the case that companies like Facebook and Google aren't using a single algorithm, but rather a hodgepodge of techniques and components (some of which are neural networks or similar, which are _trained_ using algorithms, and _use_ matrix multiplication and gradient descent algorithms or variants thereof to calculate predictions, but are not themselves "algorithms").
This hodgepodge is not solving a specific, well-defined problem. It's not guaranteed to terminate (the exact opposite actually). It doesn't produce the "correct answer" for every instance of the "problem" class. In virtually every aspect that matter it's not an algorithm. Arbitary programs are not algorithms.
While semantic debates are undoubtedly a waste of time when we use an ever evolving descriptive language like English, I couldn't help myself. Most dictionaries appear to support OP's usage, and I would argue that it is very intuitive that an alorithm made up of smaller algorithms is still, itself, an algorithm. Google's algorithm for determining whether to ban USENET groups mathematically approximates the problem and computes a solution to that problem in a finite number of steps, frequently involving repetition. Even if it didn't fit those specific requirements, however, it would still qualify under the broader definition of "algorithm" in common usage. See Merriam-Webster's definition of algorithm:
: a procedure for solving a mathematical problem (as of finding the greatest common divisor) in a finite number of steps that frequently involves repetition of an operation
broadly : a step-by-step procedure for solving a problem or accomplishing some end
"There are several search engines, with Google, Yahoo and Bing being the biggest players. Each search engine has its own proprietary computation (called an "algorithm") that ranks websites for each keyword or combination of keywords." — Julie Brinton
"… sometimes you solve a problem by coming up with an algorithm of some kind. But sometimes you solve a problem in a very ad hoc sort of way." — William H. Huggins
At the risk of sounding like I would want to die on this hill (I am certainly annoyed, but anybody can use any word to mean anything they want as long as a certain number of other people agree. I think I understand how natural language works), but this statment
> I would argue that it is very intuitive that an alorithm made up of smaller algorithms is still, itself, an algorithm
is very unconvincing. Here's an "Algorithm"
- compute the square root of 5
- for each number from 1 to n
do 5 times :
compute the
square root of 5 times
that number plus 19
- output the square root of 9 to the screen
Are you willing to accept this sequence of steps as an algorithm? I am not. And this sequence is even more of an algorithm than Google's products, at least it guarantees termination, correctly solves every instance of a (extremely artifical and meaningless) problem, and is effectively calculable (Google's and Facebook systems frequently need humans to decide or amend decisions, and the decisions are often arbitary and value-based). Neither of these things is guaranteed for Google's "algorithms".
>Most dictionaries appear to support OP's usage
I never said that this confusion isn't popular or documented in dictionaries and literature everywhere, the exact opposite in fact. I said I am annoyed that it is popular. Programs and Algorithms are different, and the difference is as meaningful and useful as the difference between "Writings" and "Poems".
Well, now I'm feeling nostalgic. This little discourse brings back memories of the good old days on comp.lang.* and elsewhere on USENET and other ancient fora. ;)
The AI/ML meaning of "algorithm" (a computationally-derived heuristic for search or filtering) is not the mathematical sense (a precise formula providing an exact solution to a given problem).
But in the sense that an opaque, poorly-understood, by either users or the Internet monopolists themselves, black box has made yet another inscruitable, unanswerable, decision which cannot be explained or described, the phrase is accurate. Over the past decade, I've observed that Google's own staff can increasingly not answer why specific actions occur. The solution seems to be to tweak the training data and/or desired results set. Given the gradual extirpation of computer-language dedicated Usenet groups, this process seems at best slow, if not utterly futile.
It's not "an arbitrary program" either, at least not strictly, as the process is itself the end-result of a specific set of steps and processes. That happen to involve machine-learning and AI.
Pointedly, the decision isn't one that's based on a human agent, review board, or process.
If you'd prefer, "because a machine-learning based AI said so" would be more accurate. But it's also more of a mouthful.
I (not the grandparent) would prefer “the heuristic”, which feels like it gives an accurate impression of how well-founded and transparently derived the decision is... But maybe it instead shows how confused I am about how, why and in which sense ML actually solves problems (tremendously so).
> comp.lang.tcl has been identified as containing spam, malware, or other malicious content.
This type of ban isn't new for comp.lang groups. In 2020 comp.lang.forth and comp.lang.lisp were banned with the same message: https://support.google.com/groups/thread/61391913/some-usene.... The response from a person with a "Gold Product Expert" badge in the support thread was that "since the Usenet newsgroups doesn't have owners there is not any way to appeal the Ban".
Cant open the link, i get the warning the link contains spam or malware. Er is gedetecteerd dat comp.lang.tcl spam, malware of andere schadelijke content bevat.
Honest question: Why cling on to Usenet in 2021? Why not use an alternative solution such as a mailing list, Discord, Element, Slack, Discourse, Vanilla Forums, Simple Machines Forum, etc? There are many other systems out there, even if you want self-hosted open source software.
I still use Usenet. I don't like mailing lists, Discord, Element, Slack, Discourse, etc. NNTP and IRC are much better.
I only started using Usenet recently (and only for text, not binaries); all of the spam messages I have seen are from Google Groups. Legitimate messages are sometimes from Google Groups too, but are also posted on other servers. (I wrote my own NNTP client software actually)
I also run my own NNTP server, for the discussions for my own projects and that stuff (it isn't connected with any other NNTP servers currently, although it could be done if wanted). Currently it does not include the web interface; those who are interested in it can help with such a thing. Web interface is then an alternative for the people who do not like NNTP; the web interface includes a clear link to the NNTP even if JavaScripts/CSS are disabled, so that if given the link to the web interface, you can still use the NNTP if you do prefer to use the NNTP.
(I also use 63-bit article numbers rather than 31-bit article numbers, although it is unlikely that they will ever reach high enough numbers that will exceed 31-bits, it is an available feature if it is needed.)
Its superiority has been in question since September of 1993, when it noticeably buckled under the load of a million or so (and quite probably fewer) commercial dial-up service subscribers. Previously Usenet had a maximum user base genereally of less than 1 million. The pre-AOL Usenet probably measured less than 1 million potential users, and a small fraction of that active.[1] It dragged on for another decade, more or less, but was largely no longer freely available to dial-up / residential users by the early 2000s, largely due to its inability to police against antisocial behaviour, or mount a credible political defence against claims of rampant criminal activity (at least some of it reasonably well evidenced).
As such, actually offering Usnet access became a net liability to most commercial operators.
1. Relying on Brian Reid's "USENET Readership Summary Reports", as of April, 1988, there were 381 newsgroups, 57,979 messages, an estimated 141,000 readers, of a total user population on connected hosts of 880,000. Cited in John S. Quarterman, The Matrix, 1990.
Every ISP used to offer free access. It was promoted right alongside other offerings. The shrinking accessibility and visibility is part of the problem. It's easier to kill something off if the only people who use it are people who don't have a place on the mainstream internet.
Look at the uproar over Freenode. No one would notice if IRC had been as obscure as NNTP over the last 15 years.
The problem with NNTP is a lack of incentives. ISPs have no incentives to keep the amount of data in Usenet. We need some kind of distributed system where the incentives to store and publish the data are there and are orthogonal to the interests of people reading and writing comments. It would even be a plus if it costed a tiny amount to write those comments, so as to reduce spam.
If one excludes the binary groups then the Usenet doesn't need that much storage at all.. The cost for that is negligible.
However one needs to put in some work to keep it running and given it's not exactly a mainstream service, many ISPs have decided long ago to shut down their NNTP servers.
Not sure I follow the argument. You're saying that Tcl programmers don't have a place on the mainstream internet?
Also, I'm not sure I understand your argument about NNTP and IRC. Why did NNTP became obscure and why did its accessibility and visibility shrink, when the same did not happen to IRC? Given that NNTP is distributed and open and superior to many things that are not obscure, accessible, and visible, how could such a thing have happened?
The way I remember it, back in the late 90s and early aughts every news group became useless due to cross posting spam. Sure spam was always a thing, but I remember groups I used to read became unreadable by the sheer number of binary files being posted.
IRC could avoid this problem because it is transitory. But I think the biggest reason is that usenet is unmoderated[0], while there’s always an op on an irc channel.
[0] Yes, there’re some things people try to do, but they always felt bolted on, and a bit awkward. I’m not even sure how widespread they are, because I don’t think I’ve ever seen them actually work. Then again, I’m talking about a Usenet from 25 years ago.
I think partly, it's that NNTP is harder to run than IRC. You need to (at minimum) set aside a substantial amount of storage space, how much depending on whether you carry binary groups, and how long you retain articles for.
The other thing is that people tended to rely on their ISPs to provide NNTP service, and they started tending to drop it in the late 90s and early 2000s, and there were few/no free alternatives at the time.
Yep. Compare to IRC which has always had an abundance of free servers and free clients that run on anything. You can telnet into an IRC server if you type fast enough. It's still the go-to support channel for major open source projects because of that accessibility.
Google and other large, centralized providers will always invariably end up having to censor profusely, if for no other reason than they want to retain their customers (spam is distracting and annoying), but also because any high profile hosting/message relaying/republishing platform that makes URLs available to the whole world is going to have a PR problem when a journo links to something unpopular/objectionable hosted on the domain with the standard "Why is Google supporting/promoting this trash?"
YouTube recently banned accounts and censored videos that contradicted the WHO - the same WHO that said you shouldn't wear masks during the pandemic.
Centralized systems are censorship accidents waiting to happen. It's super bad already, and we're still in peacetime. Imagine how shitty things are going to be if that changes and the nations that can exercise ultimate control these vast centralized systems feel that free communication via them poses an existential threat due to crisis.
Then there's the whole Instagram "no nudity" thing. Even outside of times of crisis, the de facto photo sharing default has excluded an entire category of human art and creativity. No single organization should have this much dictatorial control over our society's ability to express its own culture, nevermind its ability to permanently read (and share, and mine) your private conversations.
Build around centralized systems by running things yourself. Run your own mailing lists so your groups can communicate outside of algorithmically reorganized timelines designed to benefit advertisers. Self-host Discourse (forum) or Mattermost (web chat) or Gitea (git web). Make local backups of your servers and vpses data partitions so that you can fire them back up somewhere else when the time comes.
Your community will thank you in the long run, and you will be contributing to making all of society more resilient and more expressive/free.
The sentiment in your comment is frequently expressed on HN - censorship isn't good, we need to decentralise to escape censorship.
Ironic that this sentiment is expressed on HN, a centrally controlled site with strict rules around what can and cannot be said. "But all those rules (https://news.ycombinator.com/newsguidelines.html) are good". Well, I agree. These rules are good. They are enforced well. It does foster a decent community with interesting discussion. But let's not mistake a well run, controlled system for a laissez-faire one. HN without moderation would not be the same site, and I'd bet most of us wouldn't want to be a part of it.
This particular comment is strange though - you rail against Instagram's no nudity policy, something that almost all people agree that the site is better off without. And yet, have you tried posting nudity on HN? Let us know if it stays up. HN's control over topics that can be discussed is so firm that even potentially interesting threads like the first one around the European Super League (https://news.ycombinator.com/item?id=26854970) were shut down.
>Ironic that this sentiment is expressed on HN, a centrally controlled site with strict rules around what can and cannot be said.
There's a big difference between a general photo-sharing platform like IG and a small, somewhat topic-focused forum like this.
It's like saying a recipe submission site is "censoring" nudity posts. I guess that's technically true in a sense, but it's also not really what the site's for.
HN itself is an independent site, as opposed to something like a Facebook group. The strict rules are ones we all agree to by joining. So HN by nature falls in the category of "running stuff yourself" since it's not part of a giant centralized conglomerate like Google or FB.
> Ironic that this sentiment is expressed on HN, a centrally controlled site with strict rules around what can and cannot be said.
I say this on my own website, on other sites, and here too - it's not an exclusive thing. This isn't the gotcha you seem to think it is.
> you rail against Instagram's no nudity policy, something that almost all people agree that the site is better off without.
Citation needed. This is your opinion, not fact. There are many thousands of fine art photographers who can't share the majority of their work on IG, and hobbyists who have lost their entire accounts and audience simply because Facebook's censors can't tell a male nipple from a female one. Many people into fetish clothing and jewelry design can't even post non-nude images showcasing their products. There isn't the wide consensus you imagine.
> And yet, have you tried posting nudity on HN? Let us know if it stays up. HN's control over topics that can be discussed is so firm that even potentially interesting threads like...
I actually wrote and operate a little hobby site that exists solely to detect and highlight stories upvoted by HN's readership and marked as off topic (and hidden) by HN's administrators: https://orangesite.sneak.cloud
People should be able to decide for themselves what they see and read, including delegating that authority. HN, for example, has the "showdead" preference option for opting in to seeing comments marked dead by the mods.
HN strikes a good balance, and lets people opt in or out of comment moderation. Facebook and Google do not.
> People should be able to decide for themselves what they see and read, including delegating that authority. HN, for example, has the "showdead" preference option for opting in to seeing comments marked dead by the mods.
> HN strikes a good balance, and lets people opt in or out of comment moderation.
While this is a good idea (and I use the showdead option myself), it is incomplete. To complete it might include such thing as:
- Add another preference option to force everything to be displayed in chronological order.
- If you have enough disk space to allow it, allow replies to dead and old messages, which are automatically also dead.
Hacker News isn't big enough to be a problem. It's not big enough to steer the global conversation. Censoring something from HN doesn't mean that huge populations never hear about it.
Not being able to cope with nudity seems to be a USA specific phenomenon. I female friend of mine was almost arrested in the US for dancing too suggestively, and she was far from being nude. Seems to be a puritanism thing.
Instagram is a US company, staffed primarily by people from the US culture, so the problems with this USA-specific phenomenon are being exported to billions of others who aren't part of that culture.
I know. Its sad. And it happens with other things too. That is the price we (the rest of the west) pay for never having emanicpated ourselves. This "the US defines what west is" thing pretty much goes on my nerves since I realized how it works. It is also likely a reason for anti-US-attitudes. The loud and extroverted bully is never really a darling of everyone.
> so the problems with this USA-specific phenomenon are being exported to billions of others who aren't part of that culture
This is the case on the vast majority of technology developments -- things to solve mainly US centric problems are pushed out with massive amount of money to try to "solve" problems that don't exist.
A byproduct of this push is often to create those problems in other cultures. It's like the British empire of the 17th century.
It's not US culture that's the issue - ever heard about the utterly ridiculous regulations Germany has on pornography (tl;dr: a simple "are you 18" prompt isn't sufficient, you need a certified online identity provider verifying your age against a government issued ID)?
Reddit, in contrast to Instagram, hosts porn of all varieties - from softcore nude art to hardcore fetish stuff.
The issue with the US is "family friendly" brand/advertising culture - no big brand advertisers want their ads to go next to a blowjob - and credit card regulations.
You're just objectively wrong. To the extent that _anyone_ cares about Usenet for actual discussion purposes, there are lots of other providers you can choose from.
