ihnp4!{cithep,wlbr!callan}!tim
ihnp4 was very well connected, so it was reasonable for me to expect that you could find a way to get the email to it.
The rest of my email address is saying that ihnp4 should deliver the mail to cithep (Caltech High Energy Physics) or wlbr (a machine at the Westlake Village, CA offices of Bunker Ramo). If it goes to wlbr, wlbr should send it to callan (a machine at Callan Data Systems). I had accounts on cithep and callan, both with user name tim.
You didn't have to start at ihnp4, of course. As long as you could find a way to get it to one of {ihnp4, cithep, wlbr, callan} you could get it to me.
Periodically there was a posting on Usenet giving a list of machines that would publicly relay mail and their interconnections. (I want to say this was monthly and in net.general, but I'm not sure).
Given that and a list of machines that your site connected to, you could find a path from your site to me. Let's say you were in Boston, and you knew your company exchanged mail with decvax (a machine at Digital Equipment Corporation). You could see from the connection list that decvax connected to trwrb (a machine at TRW Redondo Beach, CA, I believe), and trwrb connected to wlbr, so from your machine the email address decvax!trwrb!wlbr!callan!tim would work to reach me.
Usenet is still around, I still read it most days, with trn.
Most newsgroups are either dead or pure spam, but some are active with real content.
It didn't seem to count towards our download limit either. Not sure why not, but that was awesome.
I think this is what really did usenet in for me. I used to use newsgroups quite a bit in the early 2000s, then moved on to forums, and then of course social media took over (and now I'm substantially over that).
Given that usenet is now past its popular heyday, I thought it would also be past the peak of spam - and it probably is - but having just registered with eternal-september.org, I'm still pretty surprised by how much spam is in some of these groups. Granted a lot of it is old (because I told Thunderbird to just download all the headers), but there's loads of more recent stuff as well: e.g., crap about COVID in all kinds of groups, as if we're not hearing and reading enough about it already.
This is kind of a shame. If newsgroups had a viable mechanism for combating spam I think they'd still be a great place to hang out because they were and are a one-stop shop for such a wide range of activities and interests. In addition, and ignoring the spam, the reading and writing experience is still really pleasant.
What makes it stand out more is that a bunch of the newsgroups I've subscribed to are very inactive, it turns out. I just yanked in anything related to retro computing and gaming and am slowing pruning the list back to include only active groups.
The more active groups have spam but they also have self-promotional material that might or might not classify as spam, and I'm seeing a few mass crossposts as well.
Overall it's not as bad as I remember it being the last time I looked at getting back into usenet, which was probably 8 or 10 years ago. I think I perhaps gained a slightly skewed impression as the first few groups I went through were all pretty spammy.
I loved USENET. Threaded discussion is intoxicating. I learned and shared a lot.
Came back after the spam waves to visit some retro computing groups. A few are active. Cool. Low spam.
> ...the executives of all these corporations who made the decision to fund these n*s all deserve to die in mass-shootings.
The basic concept is simply copying and forwarding email-like messages, grouped into a hierarchical set of newsgroup feeds. A host would subscribe to other hosts, and would allow other hosts to subscribe to their own feeds. And then just push/pull this through the network.
Originally, this was done all with shell scripts dialing up remote hosts by modem and copying compressed datasets nightly. This developed into a full suite of protocols called UUCP. Later, Internet protocols like the Network News Transfer Protocol started carrying some, then all, of the traffic.
In principle, it is unregistered, unpaid, and decentralized. You start exchanging with someone else who has the newsgroups you want to send and receive.
In practice, both in the past and now, you subscribe, usually commercially, to one of the services with the bandwidth and storage to host a reasonable portion of Usenet. ISPs often provided NNTP Usenet service into the early 2000s, along with SMTP/POP for email.
Back when UUCP was still around there wasn't much binary data in groups. It took a while to develop algorithms to encode binary data (uuencode, base64), the protocol was text only.
Your disk, however, would only have the messages you'd specifically requested and pulled down over your modem. (Or, if you went way back to the vt100 days, "your disk" was the central server which had it all.)
My 1M message Usenet archive is 5.6GiB of space, and I believe this covers ~30 days of non-alt.binaries messages. The data is a decade old, so I don't recall the exact details of collection anymore.
One can dial into your ISP, sync News (fetch new messages in groups one is subscribed to and send messages) and then go offline and read/answer the messages without paying per minute and without blocking ports for other dial-in customers.
This also worked between different news servers in distinct networks with unreliable connections.
For the how paid: In some regions it was part of the things a "good" ISP had to offer their clients. Also back in the days™ large parts of the internet was run on machines in some university labs.
I still don't quite get the user experience. What if you're subscribed to alt.binaries? Do you then sync all the binaries when you go online? That wouldn't make sense. Perhaps it only synchronizes on a level below that, alt.binaries.something.something? But that may still be too much.
And what if my ISP's server didn't have the newsgroup I was interested in? Is it like DNS, where it would then fetch it from somewhere else and then cache it? Or did most servers store the entire data? Or an ISP would just store the messages that at least one of their customers subscribe to?
For what you download in sync depends on your config. alt.binaries is special, I believe relatively young and many providers didn't offer them to begin with ...
When fetching you configured your client "please fetch all of comp.lang.cpp" and it would do. I believe you could also say "just give me the headers" and load only selected messages on next sync (i.e. for fetching only binaries from messages with interesting subjects)
If your ISP didn't have a server, there were multiple open ones in universities or other ISPs (not all locked down for their customers only but either didn't care or saw it as an ad for them), but they could be far away (modern cross ocean cables are younger ...) and busy/slow.
And yes, a server can decide which groups it syncs.
If you were one of the fortunate ones to have a hard-wired connection, well, that happened a lot faster.
Newsreaders were awesome for reading newsgroups. You were able to add someone or a topic to a kill file for one group or all groups.
As binaries got bigger and bigger, the retention time for them on the remote system got shorter and shorter. You might have 7 days of binaries and 90 days of everything else.
For a while some forum software would be accessible via NNTP which was awesome. It let more people use the forums but gave more experienced users much better tools.
Typically you synced the equivalent of headers and bodies separately (for the groups you were interested in). This way you only grabbed what you wanted. For binary groups, people settled on patterns in subjects, etc so that clients could automate downloading of files (binaries had to be encoded and split into parts due to post length limits).
On regular text type newsgroups you might set your client to download headers+bodies in one shot for offline mode. With offline mode you would then just download what's new vs your cache. Then work against your local cache for reading/replying. Once online again, your batched replies could be sent.
If your ISP didn't have the newsgroup you wanted, you would have to point to another NTTP server. For a while there were public ones that anyone could use. Over time, those public ones went away or dropped problematic areas like alt.binaries.*. These days Usenet is available as a for-pay service.
This was also where the term “spam” really got popularized (see https://en.wikipedia.org/wiki/Laurence_Canter_and_Martha_Sie...) so many things which we now take for granted were being tried at scale for the first time there. There was a [relatively weak] moderation system which could be used to restrict posting and there were bots which would retroactively cancel messages so you might not see so much junk if a group used such a bot and ruthlessly nuked spam.
Regarding ISPs, in many cases it worked like your last case: start storing anything that at least one client has asked for recently. This was a notorious source of disk space consumption if you ran a low-capacity server and then had someone fire up a client, request the world, and then stop reading it but your system was still downloading everything.
That's one of the reasons ISPs stopped offering them - they'd have to have the infrastructure for all of their customers to support something that perhaps only one person would actually read.
Universities would often not carry the complete alt.* heirachy for example (especially in later, post 2000 years where the binary groups started getting big).
Yes. Your news reader (nee aggregator) would use the NNTP protocol to tell the server "please give me a list of articles in groups X, Y, and Z". Your reader would know you read up to article 234 in a group previously, but now there were 266, so there were 32 new ones. You'd tell the reader UI to go into that group, which would cause it to download the new articles (headers) and display them to you. If you saw a particular thread you liked you'd select that and the body of the article(s) would be downloaded.
If you wanted to do offline reading there was a file format call SOUP which would allow you to download it to your local machine (remember, this was often in the age of dial-up). There is no article on SOUP, but a similar idea was QWK for BBSes:
* https://en.wikipedia.org/wiki/QWK_(file_format)
To distribute articles/posts globally an NNTP server would talk to other NNTP servers: technically peer-to-peer was/is possible, but it was often hierarchical. One could also have site/local-only groups and global groups.
Messages were transferred overnight by copying the latest changes between each uni, you could post to the use group of choice using a usenet client. I was very excited to have access to it at the time, it was rare to have a connection outside of a uni.
There were source code feeds to where you could download the latest sources and try to compile them yourself, I remember fiddling with things like flex and bison.
It was full nerd - pretty well everyone was an academic, student or worked at a uni or research lab, you'd have contact in almost real time with anyone - this was a new thing.
You'd interface with it using a glass tty - I had a vt52 which connected to a Sun 3 box, which was our development computer. I remember getting a 300 baud modem and connecting to it at night from home with my PC, or maybe it was my amstrad still, PC's were very expensive still. It was a fun time.
Edit: just remembered we only had one dial in modem to the uni, and I'd use it to dial in from home, I remember we used to fight over access to it - eventually they got a second one because the sys admin was sick of not being able to get access to the network at night. They to were very expensive, I think they were 2400 or maybe even 4800 bps - a crazy speed then.
What it was is basically a massive collection of discussion groups, organised hierarchically, hosted on local machines that exchanged these messages with other machines. There was a variety of newsreader programs available that you could use to read and post messages on these discussion groups.
Not every usenet host supports every group. There were a couple of top-level hierarchies, like rec.* for everything recreational, like hobbies and games (I was an avid reader of the rec.games.frp.* newsgroups, which were about roleplaying games); sci.* for everything science, comp.* for everything computer related (some of which was admittedly also recreational). There was also the alt.* hierarchy where literally anything goes. Many hosts only supported a limited selection of the alt.* newsgroups.
You'd post under your own email address, and most newsreaders also supported emailing people directly instead of responding in the group.
I believe there was a big revamp at the end of the 1980s where everything about usenet got changed. I only know the situation from after that reorganisation.
If you are an Emacs user, Gnus is another way to get the old school experience.
Another text one that was popular is tin, which I've never used but which, like slrn, is still maintained and updated.
I still use Thunderbird most often.
For anyone who doesn't want to sign up to try it, there's also:
news.aioe.org
https://archive.org/details/matrixcomputerne0000quar/mode/2u...
I lifted a copy off a friend who was thinning his book herd a decade or so back. The first edition, published in 1990, predates the commercial Internet, World Wide Web, Linux, and all of what's transpired since, though the introduction pressages much by noting the role of existing networks in the the then-fresh events of April 15 – June 4, 1989, known as the six-four incident, and culminating in the Tiananmen Square Massacre.
The book has a 15 page section describing Usenet, including user and traffic statistics from 1979--1988, mostly compiled by Gene Spafford and Brian Reid. Usenet is only one of numerous conferencing systems covered, others being BITNET, UUCP, FidoNET, VNET, and EASYnet
For April, 1988: 381 newsgroups, 57,979 messages, 130.949 MB, 4,152--7,810 hosts (depending on who counted and how), and an estimated 141,000 readers (people) from a total online population of 880,000.
Larry Wall, working at JPL, is named ... as the author of rn.
Much of the book, though, simply discusses the many many individual computer networks. The notion of a single global end-to-end interconnected inter-network was still very much a novelty.
In one section, the fact that the principle public interconnect between North America and Europe was a single 9600 Baud link is mentioned....
Originally, Usenet used UUCP for transport; later on NNTP was defined so Usenet could be shared via TCP/IP. In the early days, only DARPA contractors (which included many university computer science departments) could be on the Internet; second-class citizens could use UUCP and modems.
In the early there were only a few dozen to few hundred computers, mainly in universities and computer companies. You list all the networked computers in a single file called /etc/hosts.
In the late 1980s universal protocols and funding from Senator Al Gores 1991 Information Superhighway legislation allowed the computers anywhere in the world to interconnect.
My usenet email address from 1976 still works.
Jon Udell's 1999 book "Practical Internet Groupware" was an excellent resource on how to set up this sort of infrastructure by gluing together various protocols:
An NNTP server had a list of discussion forums: clients connected and got the list, and the user could selectively subscribe to the one's they were interested in. The client would keep track of which messages were read, and so when new ones were posted you'd only see new messages by default.
Each NNTP server could talk to other NNTP servers. They would exchange a list of forums each had, and then the articles had in each forum. Each article/post had a unique ID (basically like the Message-ID in e-mails), and so each could request articles from the other that they did not already have. New messages could be posted to either server and they would 'cross pollinate' (multi-master basically).
Each NNTP server could technically talk to multiple other NNTP servers, so cross-pollination could be very wide, but in the early days with low-kbps modems and expensive long distance, there was a lot of local hierarchical hubs.
There were a few key hubs that had fancy long-distance T1 and such that could transfer across continents and oceans as a bit of a service for the community.
One would "post" a message and periodically one or more systems would poll your system. Either over the ARPAnet or via an actual modem. It would fetch all of the posts that had been created on that system and add them to its list of new posts, and it would send you all of the posts you hadn't yet seen.
In a directory on your system would be all of the posts from everyone on Usenet for all of the groups you cared to "follow."
Back in the 80's I could read everything posted in a couple of hours in the morning. :-) That changed quickly.
Slight difference -- there would be no main "front page," -- only individual self-hosted hubs that could choose which "subreddits" (newsgroups) they would carry.
Typically the posting news server (i.e. your own news server) was responsible for validating that you used your real email address.
Originally there was no validation, but as spam rose validation became required.
News servers required you to login, originally they were open to anyone.
That meant that unlike with Reddit, an idiotic interface redesign only ruined it for users of one client, and only until they got time to switch to another client.