John Gilmore

Giganews Usenet History > John Gilmore

John Gilmore, along with Brian Reid and Gordon Moffett, is credited with founding the Usenet alt.* hierarchy. Gilmore was among the first employees at Sun Microsystems, and was one of the founders of Cygnus Support, started in 1989. Gilmore has also contributed to various free software initiatives, including the GNU project, as well as founding the Electronic Frontier Foundation and the Cypherpunks mailing list and co-authoring the 1985 Bootstrap Protocol.

Gilmore is an avid political and social activist with interests in United States security policy, freedom of information, drug policy reform, and the One Laptop Per Child project.

Interview (4/5/2007) with John Gilmore:

1. What benefits has Usenet provided your professional or academic life?

I first learned much of what I know about Unix and C from Usenet.  It gave me access to a huge community of knowledgeable people.  Either I could just watch, like a fly on the wall, while they discussed things interesting to them. Or I could ask my own questions or start discussions. I’d been in computers for a decade before joining Sun, but had no experience with Unix; Usenet was how I ramped up quickly, without pestering my more experienced colleagues with too many ignorant questions.

2. Why do you think the Renaming was originally met with such resistance?

I don’t know. I stopped using Usenet a decade ago. The newsgroup names were fairly predictable when I left, but I haven’t seen how they have evolved since then.

3. Usenet II was an attempt to provide the old structure of Usenet, including the old newsgroup naming schemes and a cabal-like steering committee. By most accounts, it was a failure. What was your opinion of the Usenet II project? Why do you think it failed?

Usenet II?

4. Are you surprised at the size of the alt.* hierarchy, even considering its very open newsgroup creation policy?

I’m not surprised. We designed “alt” as an escape hatch from the restraints imposed on the other newsgroups. I think that if the rest of the newsgroup administration had been more open and inviting, alt would have had a lot less traction. By the time alt.sex was created (and never got a home elsewhere), it was clear there were going to be scores of alt newsgroups, just under alt.sex.

5. Do you actively participate in any newsgroups? Which ones?

I’m out of Usenet now. I have enough trouble keeping up with email, paper mail, personal projects, work projects, a few IRC channels, and friendships.

See, the Usenet was a much more approachable place when fewer people used it. It was less of a firehose and more of a community.  I’m really proud to have been among the people who helped it grow and evolve — both in the software and in the social aspects and connections — but it’s like my old childhood home town. I don’t live there any more.

6. Has your technical or social experience with Usenet been beneficial to your other endeavors (the EFF, the GNU project, etc.)?

Sure. The Usenet experience is all threaded through my other work.

For example, the quote I seem to be most famous for, “The net treats censorship as damage and routes around it”, came directly out of my Usenet experience.  I was actually talking about the Usenet when I first said it. And that’s how the Usenet works — if you have three news feeds coming in, and one of those feeds censors the material it handles, the censored info automatically comes in from the other two.

The same thing happens if the first feed dropped some articles because it ran out of disk space, or had a power failure. Damage and censorship look just the same.

It turns out that one of the thought-experiments that eventually led to the Internet used similar strategies, to design a network that would readily survive a nuclear or conventional attack on most of its nodes or communications links. See Paul Baran’s seminal 1964 work at RAND, posted here:

https://www.rand.org/pubs/research_memoranda/RM3420/index.html

As another example of what I learned from Usenet, I knew on a theoretical level about the low emotional bandwidth of email and Usenet articles, but it was really brought home to me by Brian Reid and other SF Bay Area Usenet administrators. They organized regular physical meetings among people running Usenet nodes. Not only did we figure out how to more cheaply and rapidly route articles among our computers — we also met each other, shared food and drink, built friendships, collaborated on other things, vented and improved ideas.

We built a human community, not just a computer network. (The annual USENIX conferences always included an all-evening “Usenet BOF”, which was a good start, but didn’t offer enough time to really get to know the larger crowd of Usenet admins.)

Everyone who uses computer communications — or even phones — needs to relearn the lessons of physical human contact once in a while.

Later, when I was co-leading Cygnus Support, we hired a lot of “remote employees” who worked in their own homes from various places around the country and the world. We were able to hire some really top engineers, by giving them freedom to live anywhere they wanted, and work whatever hours they wanted.  They gave us their hearts and minds, and their loyalty, and their hard work in return. But we made sure to schedule and pay for a lot of physical meetings and long visits, to build the human connections among the centralized and distributed employees.

Later, for a time, I ran “BayFF” meetings in San Francisco, to provide a physical place where EFF members could congregate, meet, learn, and become friends. Most of our interactions were in cyberspace, but our friendships and collaborations had roots in the real world.

And, of course, the Usenet *was* the support community for the GNU project, before ordinary people could get on the Internet. Before students at ordinary colleges could get on the Internet. We wouldn’t be where we are today with free software, had we not collectively built a supportive community all over the world via the Usenet.  Not only were questions answered and new released posted, but large numbers of bug reports, patches, and new contributions went through Usenet.  I still have archives of some of them. Larry Wall wrote the “patch” program specifically to install patches that had been posted to the Usenet — and of course, he posted it to the Usenet! “Sources” groups like comp.sources.unix (mod.sources) or net.sources were where many free software programs were posted in their entirety.  I strongly suspect that most of the people who did computer science in the 1980s and early 1990s were first exposed to free software and the GNU project via the Usenet.

FAQs were also invented on the Usenet. It was the Usenet that taught our community to think about what questions ordinary people might ask (over and over). Skill at predicting those questions and writing the answers to them was rewarded many times over — by not having to wade through hundreds of newbie questions in your favorite newsgroup.  It’s sort of like the Zen concept of cultivating “beginner mind”.  The knack of looking at my projects from that “outside” point of view has helped me a lot since then. Every announcement I write or edit gets a critical eye — what would the man on the street say if I read this to him or her? “What’s a DRM?”  Guess I’d better rewrite to make that clear…

7. How did you feel Usenet would evolve with the advent of the World Wide Web?

By the time the Web caught on, I was deep in the startup years at Cygnus Support, so I had very little time for Usenet. I wasn’t giving it much thought.

8. What role do you think Usenet had in the evolution of web forums and blogs?

BBS systems existed before Usenet, but they generally had trouble attracting more than a local audience. Usenet really gave us access to some of the best-of-the-best in many topic areas.  Web forums really drew their roots from BBS forums — you have to go to a particular place controlled by one guy or company, log in, and that place controls what can be said, what can be censored, what’s on topic, what’s not.

Blogs may have drawn a bit from moderated newsgroups, in the sense that e.g. rec.humor.funny was a reflection of the taste and humor of one person. Also, like a moderated newsgroup, many of the most popular blogs don’t contain much original material — they mostly point out interesting things contributed by other writers.  The chronological structure of blogs also has strong parallels in the Usenet. The old stuff just falls off the bottom, and when you pop in, you’re always seeing new things.