USENIX '99: That Dinner in Monterey

I was talking with an AI today about the USENIX Annual Technical Conference in Monterey, June 1999, and the conversation shook loose a memory I hadn't revisited in a long time. I want to get it down before it drifts further.

The Terminal Room

If you never attended a USENIX conference in the nineties, the terminal room was the spiritual center of the whole event. A room full of workstations and high-bandwidth network connections (a T1 was exotic infrastructure back then), running all night. Engineers would drift in after sessions, claim a machine, and just work. Or argue. Or both.

At USENIX '99, the terminal room crowd included Eric Allman (who wrote sendmail and syslog), Kirk McKusick (BSD filesystem, Berkeley legend), and Miguel de Icaza, who was building GNOME at the time. Charles Hannum, Perry Metzger, and Christos Zoulas were around as well, in and out of the terminal room and at the pub before or after dinner.

The conference that year had both the FREENIX track and the Extreme Linux Workshop, which explains the specific density of kernel developers and BSD people in one place. I was there as a FreeBSD and Solaris engineer, which was a pretty comfortable position to be in that particular crowd.

Evi Nemeth was there too. She was selling t-shirts and I bought one from her. This would have been roughly a decade before she disappeared at sea in 2013. She was warm and funny and clearly beloved by everyone who stopped to talk to her.

The Dinner

It was June in Monterey, which means the sun was still up when we sat down. A large group, more than twenty people, ended up at a restaurant or pub with outdoor seating, long picnic-style communal tables. I know for certain who was there: Linus Torvalds, Jordan Hubbard (FreeBSD co-founder, that year's FREENIX Program Chair), Wilfredo Sanchez (who was doing Darwin work at Apple, building the BSD core of what would become OS X), Christos Zoulas, Adam Prato (a serious systems engineer who ended up at Two Sigma), and me. There were others I'm less certain about after twenty-five years.

I remember Zoulas in particular because he'd gotten NetBSD pkgsrc running on Solaris and was calling it "Zoularis," which delighted me given my day job. That kind of thing was exactly the conversation at that table: technically specific, slightly ridiculous, done for the love of it.

What I remember about the dinner overall is that the BSD people and the Linux people were getting along fine. By 1999 it was clear which way the server room was going, but nobody was being a bad sport about it. The arguments were about kernel architecture and portability, not territory. Linus was quieter than you might expect. Jordan was direct and confident. I knew who Wilfredo was before that dinner because I owned a NeXTstation and had been paying close attention to where NeXTSTEP was going after Apple acquired NeXT. Meeting him felt like a specific kind of lucky. There were at least two Toshiba Libretto palmtops running FreeBSD on the table, because of course there were.

Why Write This Down

There's no public record of that dinner. No photos (nobody had a camera phone, and digital cameras were novelties). No tweets. If anyone wrote about it at the time, it was on a mailing list or IRC channel that's gone now.

The only record is in the heads of the people who were there. And heads are unreliable archives.

That seems worth writing down.