News for hackers. Stuff that matters.
Re:Usenet – Let's Return to Public Spaces by iso1631 on Thursday February 13, 2020 @08:40AM (#22317619)
Usenet never survived Eternal September

https://en.wikipedia.org/wiki/Eternal_September

Re:Usenet – Let's Return to Public Spaces by crankylinuxuser on Thursday February 13, 2020 @09:16AM (#22317850)
No, it never survived the late 90's campaigns against the RIAA trying to stop piracy at the ISP level. Thats when all the big ISPs yanked support, especially after alt.binaries was sharing damn near everything.

And they were big bandwidth sinks due to the distributed model usenet has, so it was easy for ISPs to yank the carpet and look at the immediate gains.

Re:Usenet – Let's Return to Public Spaces by kevin_thibedeau on Thursday February 13, 2020 @10:30AM (#22318436)
Usenet wasn't dependent on binaries groups in the 90's. There was plenty of active discourse. Non-commercial hosts could easily drop binary groups. Commercial providers were the ones stuck in a bind.

The massive cross-posting of troll threads in the late 90's is what helped kill it off. The lack of effective moderation controls is the biggest weakness of Usenet.

Re:Usenet – Let's Return to Public Spaces by dspillett on Thursday February 13, 2020 @10:30AM (#22318434)
> RIAA ... especially after alt.binaries was sharing damn near everything

While this didn't help, I don't think it was a key factor for ISPs. Most ISP NNTP services didn't carry binary groups anyway for bandwidth cost reasons. There were some issues with groups that linked to where to find copyright covered material, but the RIAA and their ilk were going more for the direct sources at the time. The public list/pointer resources were actually useful to them as lists of places to chase down.

Of course this led to people paying for external NNTP services which did carry the binaries groups. These services were obvious targets for the RIAA and other such groups unlike the ISPs.

> they were big bandwidth sinks due to the distributed model usenet has

This was a far more important reason, even without carrying binaries groups Usenet could consume a large amount of bandwidth. As well as the incoming load, and the bandwidth used sending data to clients, back then modem access was common and NNTP lead to people leaving the line open to download a huge pile of stuff (most of which they'd discard without reading anyway) meaning ISPs would have to invest in more modem racks, impose unpopular limits, or be perpetually busy, any of which would lose them custom.

Another significant issue was the cost of building and maintaining the servers required too. To run an NNTP service for a noticeable amount of users with reasonable performance you needed an arrangement with impressive IO performance stats for the time, and the access patterns (constant & random) could be murderous to the drives, sometimes chewing through them as fast as they could be replaced.

Re:Usenet – Let's Return to Public Spaces by ajross on Thursday February 13, 2020 @09:57AM (#22318159)
Yeah, exactly. I worked for a small employer in the early mid-90's who maintained a NNTP feed (and had for years), and ultimately had to shut it down because it had basically become a giant unmanageable firehose of porn and piracy. The actual human-written text content had dropped down to the 1% level or worse.

Usenet died for the same reason Facebook won: people exploit "free" forums in ways that ruin the experience. It's simply not possible to have an unmoderated discussion environment in an unrestricted internet, which is why we're having this discussion on a moderated site.

Re:Usenet – Let's Return to Public Spaces by pwg on Thursday February 13, 2020 @03:07PM (#22320985)
I don't think the RIAA's efforts had much effect on Usenet, since in that time-frame the RIAA all but ignored Usenet. They were much more in shock-and-awe over Gnutella (https://en.wikipedia.org/wiki/Gnutella) at the time.

What I think led to the decline was at least a combination of:

1) an overrun of SPAM - Usenet was first to be hit by the spam flood, email SPAM came later as Usenet usage died off.

2) the 'newness' factor and the 'new shiny object' factor of the web drew away existing users, and resulted in new users never joining (i.e., pictures, fancy formatting, etc. all made plain text character Usenet posts look "old-fashioned"). And once you have a situation where new users don't join, and some number of existing users continue to depart, you are on a downward slope to disappearance.

Both of the above helped contribute to ISP's dropping NNTP service. Bandwidth costs for #1 (plus bandwidth if they were carrying alt.binaries.*) and a drop in NNTP usage due to #2. They (ISPs) no longer saw offering NNTP as a sales factor for obtaining subscribers, and once NNTP was no longer a "hook" to bring in subscribers, it was only a matter of time before they decided to just drop it entirely.

And of course ISP's dropping NNTP accelerated the issues around #2 above.

Re:Usenet – Let's Return to Public Spaces by u801e on Thursday February 13, 2020 @07:10PM (#22322690)
Another reason was due to the actions of former New York attorney general Andrew Cuomo in 2008 where he got several major ISPs to block access to Usenet newsgroups that were used to distribute child pornography[1]. IIRC, those ISPs discontinued Usenet service not too long after that.

[1] https://www.cnet.com/news/n-y-attorney-general-forces-isps-t...

Re:Usenet – Let's Return to Public Spaces by u801e on Thursday February 13, 2020 @10:34AM (#22318485)
There are a number of free text only usenet providers out there. Why couldn't the ISPs simply stop carrying the binary groups instead of discontinuing their usenet service entirely?
Re:Usenet – Let's Return to Public Spaces by chungus_khan on Thursday February 13, 2020 @10:01AM (#22318197)
My ISP still includes Usenet, which is pretty cool.
Re:Usenet – Let's Return to Public Spaces by Melting_Harps on Thursday February 13, 2020 @02:00PM (#22320345)
Who do you use? I had Cox because they were the only one left offering usenet (albeit capped).

I was a really late adopter to torrents because of my background in usenet, sometime around 2010 all the alt.binaries I used to frequent went dead or were broken with little par2 support.

It was a great time back in the early 2000s, though.

If I'm honest, I don't download as much as I used to, but I do miss the niche communities based on those groups, though. They have since moved to IRC but for the most part its all gone from what I can tell.

Re:Usenet – Let's Return to Public Spaces by chungus_khan on Friday February 14, 2020 @09:57AM (#22327055)
Cogeco in my case. No cap either, which is pretty cool.

Those little niche communities on usenet were what originally got me so excited about the internet. Both participating in ones that interested me and just exploring all sorts of interesting little communities.

Re:Usenet – Let's Return to Public Spaces by ape4 on Thursday February 13, 2020 @09:40AM (#22318043)
Yeah, no binaries would be a start for a new Usenet
Re:Usenet – Let's Return to Public Spaces by crankylinuxuser on Thursday February 13, 2020 @09:58AM (#22318175)
The problem is that binaries are transferred in a non-binary manner. Base64 was the preferred way. Except it was also used to push images back and forth as well for pertinent conversations.

For any method of communication, you can transfer content that someone else tries to prevent.

Re:Usenet – Let's Return to Public Spaces by chalst on Thursday February 13, 2020 @10:10AM (#22318270)
Upper bounds on per-user bandwidth would work. Implementing this is nontrivial, but it's an essentially easier problem than spam.
Re:Usenet – Let's Return to Public Spaces by shadowgovt on Thursday February 13, 2020 @10:34AM (#22318493)
Upper bounds on per-user bandwidth would help. Though for significant messages, one still has sock-puppet accounts (or aggregates of like-minded individuals with copies of the source message pooling their resources). And, of course, upper bounds create knock-on restrictions (should the system be allowed to transmit public-domain large-volume data? A bandwidth limit blocks that "valid" use too).

It's fun to game-theory how such limits can be broken.

Re:Usenet – Let's Return to Public Spaces by kick on Thursday February 13, 2020 @10:18AM (#22318332)
Who cares about piracy? It exists over HTTP, too, and in a far greater volume than a modern USENET-equivalent could ever hope to achieve, and it's not like ISPs need to be sucked up to anymore: they've already taken their ball and went home on this matter, proverbially-speaking.
Re:Usenet – Let's Return to Public Spaces by ape4 on Friday February 14, 2020 @11:42PM (#22333321)
Re:Usenet – Let's Return to Public Spaces by shadowgovt on Thursday February 13, 2020 @10:30AM (#22318439)
Unfortunately, technically banning binaries merely means binaries will be transferred in a more spacetime-inefficient manner.

You can mathematically prove atop Shannon's theorem that if the system can transmit comprehensible text information between two users, it can transmit binaries. Worst-case scenario, the users could use the text information layer to just say 'one, zero, one, one, zero, one' at each other, etc.

(Socially banning them can certainly "work," in the same sense that social banning works in any context: pushes it underground out of the moderators' lines of sight. Depending on what you're trying to accomplish, that can be good enough).

Re:Usenet – Let's Return to Public Spaces by tomjen3 on Thursday February 13, 2020 @03:39PM (#22321221)
I have given this some consideration, and what I ended up with is super low tech: require a user registration where users can't just sign up for N accounts and limit all accounts to 10 MB a day. Too little to share any meaningful binary data and way more than a person can type in a day.
Re:Usenet – Let's Return to Public Spaces by shadowgovt on Thursday February 13, 2020 @04:04PM (#22321408)
It's a good idea. What stops a user from registering for N accounts? How do you tie one account to one human?
Re:Usenet – Let's Return to Public Spaces by tomjen3 on Thursday February 13, 2020 @04:49PM (#22321773)
In my original consideration you could only sign up with government ID -- actually the original problem was what to do prior to the day reddit banned your community, so that you could keep going.

In either case it was pretty trivial to picky bag of existing IDs, if I ever solve that for the general case, you will find out.

Re:Usenet – Let's Return to Public Spaces by cat199 on Thursday February 13, 2020 @10:15AM (#22318311)
many interesting usenet groups were long dead before that due to spam... the things you mention are more like pulling life support than the cause of death
Re:Usenet – Let's Return to Public Spaces by takeda on Thursday February 13, 2020 @10:23AM (#22318372)
Spam started when Google acquired DejaNews renamed it to Google Groups, and have Usenet access to everyone without responding to any abuse complaints.
Re:Usenet – Let's Return to Public Spaces by webmaven on Thursday February 13, 2020 @11:08AM (#22318789)
That's a bit ahistorical. The "September that never ended"[0] (aka "Eternal September") was in 1993, the Canter and Siegel Usenet spam[1] was sent the following April. The DejaNews acquisition[2] wan't until 2001.

[0] https://en.wikipedia.org/wiki/Eternal_September

[1] https://en.wikipedia.org/wiki/Laurence_Canter_and_Martha_Sie...

[2] https://en.wikipedia.org/wiki/Google_Groups

Re:Usenet – Let's Return to Public Spaces by kps on Thursday February 13, 2020 @11:03AM (#22318763)
Usenet spam started in 1994, before DejaNews existed.

https://en.wikipedia.org/wiki/Laurence_Canter_and_Martha_Sie...

Re:Usenet – Let's Return to Public Spaces by iso1631 on Thursday February 13, 2020 @11:01AM (#22318742)
Not sure, wiki puts that as 2001, and I remember being annoyed with spam in the 90s.

I think the rise of web-based forums is what killed off usenet as a general tool.

Re:Usenet – Let's Return to Public Spaces by gspr on Thursday February 13, 2020 @10:19AM (#22318344)
Is perhaps the article author's name – October First – a reference to Eternal September? :-)