alt.time-capsule.reopened

A new tool for digging into Usenet history holds much potential for researchers. But it’s also a reminder that many users thought, at the time, that it was ephemeral. (They were wrong.)

Today in Tedium: Last week, my past caught up with me. In 2026, I found myself looking at every embarrassing thing I ever posted on Usenet—bad song lyrics, jokes I thought were clever but got zero replies, an argument with someone about Taco Bell, even my attempt at a makeshift pop chart. It was an unusual feeling, not unlike the feeling you might have of looking at an old book of poetry you wrote before you really had an idea of what you were doing. I feel like Olivia Rodrigo, watching her old grade-school talent show videos go viral after she became better known for videos like this one. (This is the only way I feel like Olivia Rodrigo.) A recent tool has gotten me thinking about Usenet once again—and what it means for conversations that died decades ago to resurface far from their original context. Today’s Tedium considers looking at a Usenet archive, 30 years later. — Ernie @ Tedium

Computer Chronicles Revisited is a monthly subscription newsletter that reviews the people, products, and companies featured on the PBS series that aired from 1983 to 2002. Available for download or mailed directly to your address. .

2024

The year Google stopped adding new posts from Usenet into Google Groups. (The old ones are technically still online, but it’s clear the platform isn’t particularly well maintained. I think that team is working on ensuring that a Gemini sparkle is in every Google app.) By that time, many newsgroups had been overrun by spam, so in a sense, it was probably time. But it’s always been a tough racket to actually read the Usenet archive on Google Groups, despite the fact that it was based on an archive built by DejaNews, which offered one of the first real Usenet archival efforts.

Millions of old posts, no binaries: The data nerd who made a giant vintage database searchable

This ain’t our first Usenet rodeo. Far from it. Tedium’s talked about it a lot, actualy. Back in 2017, we reported on the disruptive role that binary file-sharing played in Usenet’s evolution. (It’s weird how little we thought about things like security and safety in the early ’90s.)

The year prior, I highlighted the tale of Usenet charthound Matt Levine. And in 2019, we published an article describing how modern day Trump whisperer Michael Wolff leveraged Usenet spam to build his early journalistic reputation.

And then there’s this 2020 post on the Eternal September, a Usenet-borne phenomenon that I think holds up pretty well, all things considered.

But the thing about these stories is that I always ran into a problem when trying to resurface the thing I was talking about. Too often, I would land on third-hand websites or have to dig incredibly hard to find a suitable result on Google Groups, the platform that was supposed to archive this content.

At some point, Google unnecessarily neutered its archive. Yes, many of the messages were still there, but it seems like the company intentionally made it difficult to search through the engine, only offering the most basic sorting abilities. Which means that, if you’re trying to (for example) figure out when the term “Eternal September” was first expressed on Usenet, you would not be able to find it without a lot of extra work. (In fact, the only way to figure it out might be by referring to a Google Groups link on Wikipedia—a link that, as of this writing, gives me a 429 error.)

Usenet-Rewind, a new service by the firm Erie Data Systems, makes it far easier to narrow in on an incredibly complex archive. I asked it to conduct that same search for the phrase “Eternal September,” oldest date first, and the very first result, the very post by John William Chambless referenced in the Wikipedia article, showed up like it was nothing.

That’s despite the fact that the archive has more than a billion messages in it—or, to be exact, 1,039,260,173 messages as of this writing—dating to 1981.

More searching of stuff I researched a while ago: An early post from chartster Matt Levine, promoting his mailing list on rec.music.misc, appeared on September 5, 1995:

Subscribe to the coolest new mailing list, Matt Levine’s Top 40. THe main chart is based on my opinion. There is also a cool chart where you get to vote for your favorite songs called Netizens’ Picks. If you would like to subscribe, e-mail me at [address redacted].-matt levine

Michael Wolff’s attempts to draw attention to his Net Guide offerings started showing up in the early weeks of 1994, first in alt.online-service and numerous other newsgroups, complete with quotes from Electronic Frontier Foundation founder John Perry Barlow and early Wired editor Louis Rossetto. The drama around his company’s posts getting mass-cancelled for spamming, which hit at the end of 1994, is shared in its entirety.

As for the binaries that reshaped Usenet? Curiously, I don’t see any groups under alt.binaries, though I see a number of groups with misspellings of the word binaries. I did find some groups that specialized in software downloads, such as alt.ibm.warez, and I spotted a note that specifically states “.” I guess what I’m saying is that you should not expect to download 35-year-old warez from Usenet-Rewind.

But that’s fine. Because honestly what’s there is pretty impressive.

So, how come this new thing has a significantly better search experience than even what Google has on offer? And how come nothing else has even gotten close? Perhaps it comes down to right time, right person. Erie Data Systems founder Craig J. Stadler says that he got into Usenet in his early 20s, but and he felt the same limitations with Usenet research that I did.

“Every resource I found wasn’t even close to what I wanted in regards to research,” he says. Unlike me, he could do something about it, as he has spent the past decade and a half working on search engine architecture using Apache Solr, a high-speed open-source search platform.

“I made it my mission to make this,” he says. Stadler is a big-data guy who has built massive databases in the past, but even by that standard, Usenet is a particularly large corpus. He apparently spent a full year acquiring data for Usenet-Rewind though a number of resources in what he calls a “no-stone-unturned” approach.

“Many, many archives and resources of dumps, backups, and donations were involved,” he says. “I also built an NNTP Usenet crawler to fill in the gaps and get all new public Usenet data.”

(This was how I learned that there’s a niche RSS reader tool out there that has been archiving every issue of Tedium on Usenet for the past seven or eight years, which, nice work. At least someone’s using Usenet exactly as it was originally intended.)

The result is one of the most in-depth Usenet archives I’ve seen online, bringing a kind of modern searchability to a service that is traditionally the domain of dedicated clients that haven’t gotten significant upgrades in years. (Plus, it makes room for the casual user. It is possible to use it for free for a limited number of searches with a new account, though if you’re using it with any seriousness, you’ll likely want to pay for a $9.99/month plan.)

As someone who digs through stuff like patent filings and old newspaper articles, it has always been kind of a disappointing gap that we’ve never had a real Usenet archive online.

This fills it. But it does cut both ways.

“You should know that Usenet, the battleground for what has become known as The War Between alt.tasteless and rec.pets.cats, is the Internet’s answer to Time Warner’s 500 channels of cable.”

— A May 1994 assessment of Usenet, according to an article in Wired discussing an epic flame war between two stylistically divergent newsgroups. Truer words have never been written.

The weird feeling that stewing in 30-year-old Usenet posts gives you

That Olivia Rodrigo example from earlier is not really a one-off. Much the opposite. It’s the kind of experience many people under the age of 35 are basically used to at this point: An internet in which their youthful memories and embarrassing moments are just a click away.

The college student who saved all of their old drinking photos on Facebook or Flickr, the person with the embarrassing Tumblr blog? It’s all still online, waiting for someone to find it. Our collective digital memory has not been particularly well-curated.

But Usenet was like a big redo for a number of large groups of people, because for decades, the archive was frustratingly incomplete and otherwise hard to reach. Yes, some of it seeped out, but digging for an old Usenet post, even if you remembered your old email address, felt like an archaeological dig.

I think what stands out to me about Usenet-Rewind is that it more or less captures all of it, warts and all. I did a search for my late ’90s email address, and found numerous entries under my name. I then did a search at my email addresses for the two colleges I went to and found a bunch more posts. And I’m not going to lie, most of those posts were not written with the assumption that they would be a part of the permanent record 25 years later.

So I had my version of the Olivia Rodrigo talent show thing, with a shaky voice and slightly unsure footing, except my voice was completely digital. I was still figuring out what I wanted to put out there, so I put everything out there. Penpal requests. Song lyrics. Attempts to extract advice on my budding career. Attempts to figure out chords for songs when I barely knew how to play guitar. My own personal rock chart, inspired by Matt Levine’s.

And all of that stuff resurfaced, rushing back and reminding me that, oh yeah, you barely knew what you were doing on the internet at this time. I’m sure I’m far from alone on this front.

Which naturally leads to some privacy questions. Do you want a service resurfacing things that you wrote 30 years ago, presuming that the content would eventually disappear from the face of the internet after a month? (Perhaps this is the reason why Google intentionally weakened the search on Google Groups? Notably, Google was once sued over Usenet posts on Google Groups, but won the case in a federal appeals court.)

For Stadler’s part, he points to the fact that he offers the ability to do DMCA takedown notices. So if Michael Wolff decided that he didn’t want people to check out his ’90s attempts at spam, he presumably could send one to Usenet-Rewind and they would take it down. (That said, a DMCA request does have some limits: “Removal from our index does not delete the content from Usenet itself,” the page states.)

But what about LLMs, which have made Reddit posts into one of their most valuable information sources over the past few years? Stadler makes clear that this data was not gathered for this purpose. “To our knowledge this data is not legally OK to use in an LLM, and we have no intention of going beyond traditional tokenized search,” he says.

That said, if you did want to look at Usenet in a programmatic way, perhaps to do sentiment analysis of different groups or academic research, the firm does offer a higher-tier plan which includes API access, allowing for deeper, more specific searches of the archives (within reasonable limits). But at $39.99 per month, you’re paying for the privilege.

The fact of the matter is, this is a corpus that goes back decades, is extremely complete, and taking old posts down might be like a game of Whack-A-Mole. Put another way: Your vintage standards of privacy don’t match your modern ones, and that means your awkward moments, your fights over Star Trek plots, your bad takes on The Simpsons, will remain on someone’s Usenet server forever.

Over the years I’ve spoken passionately about the need to keep old content alive online, to discourage companies from taking it down based on whatever whim they’re feeling on a given day.

Usenet, while never fully taken offline, has never quite been as accessible as it once was. For better or worse, Usenet-Rewind solves that in ways that other tools like UsenetArchives.com have never quite been able to. For one thing, the search actually works! (Some will point out that GigaNews also has in-depth archives, which I’ll concede, but that’s not quite built for casual users to dive into.)

The fact is, Usenet hasn’t been a truly mainstream experience for a long time. (Mozila Thunderbird, not exactly the world’s most popular email client, is the closest thing we have to a news reader that folks have heard of.) But it’s packed with information, much of it extremely niche, some of it super-useful. From a research standpoint, being able to dive into this information the way I might on Newspapers.com or Google Patents is a really exciting prospect.

Here’s an example: Back in the mid-1990s, MTV would frequently hold moderated online chats with celebrity figures on AOL and, later, IRC. (I mentioned one such chat, with the Butthole Surfers, way back in 2017.) These chats were all but forgotten after that moment, but many of the transcripts for these ended up on Usenet. A few I found after a couple minutes of digging:

1. Iggy Pop (in response to a question about whether he gets stage fright): “don’t get so personal,creepo.”
2. Mike Judge (in response to a question asking if he’s real): “If you had put a question mark at the end of that sentence I might have been able to answer”
3. Chris O’Donnell (in response to a question about what it was like to play a soldier in a war movie): “The fake blood was kinda sticky but it was better than a real bullet in your leg.”
4. Marilyn Manson (in an introduction ahead of an IRC chat): “I think people should have a better attitude towards the newer Marilyn Manson fans because it’s just one more step towards bringing or ideas into the mainstream and I’d rather see them being ‘Marilyn Manson poseurs’ than listening to Hootie & The Blowfish.”
5. John Popper (when asked for his views on Kraft Macaroni & Cheese): “I believe that a scam is being perpetrated on the American people. they set up a cute innocent young looking girl to say that there’s so much cheese involved but the cheese should precede the macaroni. But on my own investigation I have found only hollow macaroni spray painted lightly with a cheese-like sauce. I would call it Kraft lightly spray painted with a cheese-like sauce macaroni. Besides, Stouffers kicks butt”

This information required someone to care enough to copy and paste it elsewhere. It might have been forgotten otherwise. And Usenet-Rewind is full of stuff like this. It’s a treasure trove, and in the right hands, it could lead to some truly fascinating discoveries. After all, when it comes to history, ephemera is often more effective at getting an on-the-ground assessment of a culture than high-level nostalgia.

Usenet can be best thought of as a network of ephemera, and it helps fill the gaps of the pre-Web era that the Internet Archive has rarely been able to capture. (Something tells me that opposition researchers are going to love this thing, because odds are at least some modern-day politicians posted on Usenet in the ’90s.)

Can a thing be an important resource and a potential privacy violation? It’s an open question. But I’m just happy that vintage Usenet is a little easier to access these days.

--

Thanks to Craig J. Stadler of Erie Data Systems for taking the time to chat about this interesting thing he made. Give Usenet-Rewind a look! You might even find yourself hiding in one of these old newsgroups.

Find this one an interesting read? Share it with a pal!

And if you love old digital culture, our sponsor is very much gonna be up your alley.