I went to the FSF associate member meeting this week. This is the third year they've had the meeting, and my second time attending. The meeting occurs every spring when the FSF's board of directors gets together in Boston; the board members give talks about various topics relating to the FSF and to free software, and afterwards everybody eats at a restaurant called the Middle East.

Other people's accounts of this meeting: Marty

the Gates building

This year's venue was somewhat more pleasant than the last, since the room we sat in all day was a stadium-style lecture hall with desks for everyone to rest their laptops and moleskines on.

Moleskine count: at least 6. Attendees: at least 80. Girl count: Two presenting and at least four in the audience. Laptop survey: mostly Thinkpads. Camera survey: mostly Canon ELPHs. We had an excellent view, sitting way up top in the back :)

The building was MIT's Stata Center, a bizarre looking building of which a section was named after Bill Gates. We were advised to salute Mr. Gates's name appropriately as we exited.

The main features of the building's interior were the lack of right angles, and the omnipresence of shiny red surfaces. Now, I like shiny red things a lot, but toward the end of the day the color was getting on my nerves. (The lighting also seemed to get dimmer in the afternoon, but this may have just been my imagination).

actual eventage

I have 23 pages of notes in my moleskine that I took as I listened to the day's talks. (I also was obsessive enough to take photos of most of the speakers. I think John did too). What follows is basically a moleskine-aided brain dump. It omits anything I didn't find interesting, and anything I found so interesting I forgot to write it down.

If you'd like to comment on my notes or offer corrections (there are plenty of spots where I may have misheard, misinterpreted, or left out important parts) - write and tell me so I can update this page.

The talks were videotaped with intent to post on the website. Nothing has been posted yet (Marty recalls that they made the same promise last year and never posted them) but if they become available, make sure to watch Eben Moglen's talk, it was kick-ass as usual. I also particularly enjoyed Sussman's and Lessig's talks. (Marty thought Sussman's talk was "ridiculous" and Alex thought it was "awesome" and I have to agree with them both.)

Oh, and be sure to check out the Lessig/RMS cage fight death match. Will Eben Moglen convince them to hug and make up? Will RMS decapitate Lessig with a chainsaw? Will Peter Brown interrupt the fight so we can all go to dinner? If they post it, watch and find out!

the talks

Last year, the FSF's website went down the morning of the meeting. I don't know the whole story, but their sysadmins were putting out fires all day and couldn't make it to the meeting. So I was glad to see that one of them stepped up to the podium yesterday. He described the duct-tape-and-chewing-gum construction of the server room's furnishings, and mentioned that their bandwidth is about 5 TB/month.

The executive director's talk is generally a state-of-the-onion kind of thing, so we learned that the FSF now employs nine people, that the website's design has changed (a lot) since last year, and that David Turner was the first person to earn a recording from RMS for recruiting three new members.

Geoffrey Knauth was next. His talk wasn't flashy, but I really enjoyed it. He started by writing "FREEDOM" on the board, and then "LAW". These are the things that the FSF has covered, that they do well. With the audience's help, he covered the board with other terms describing areas we need to work on and people to reach out to. (you can see the finished board in the picture).

We need to get some economists on our side, understanding that free software is economically viable. ("We don't want to be labeled as communists, because we're not communists! We're people!") Somebody mentioned the idea of preparing an information packet for economics professors ("YES!" he said, "With lots of B.S.!"), and even a packet for economics students on how to talk to their professors about free software. Someone else mentioned that in his company's switch to free software, something that got scribbled down as "strategic flexibility" was an important consideration, in addition to TCO. I also wrote down something about getting away from "the development curve of proprietary software".

He was also concerned about the software used in education. Kids learn on Macs or Windows, depending on the choice of one or two influential people at their school, and aren't really exposed to free software. He pointed out the scenario of graduate students who run GNU/Linux on their own but are forced to work on Windows to do their research. Teachers are trained in the use of certain educational software (Geometer's Sketchbook is one example; KSEG is the possibly inadequate free alternative).

He mentioned rumors from a friend of his who works in the government: that 30% of government servers run GNU/Linux, and that two major things the government would want from free software are FIPS-140 compliance (whatever that is), and piles of documentation.

Something that was mentioned several times yesterday was marketing: getting the word out, and making sure it's positive. I think it was around this time that Mitch (a fellow Ithacan) mentioned the Ithaca free software meetup. This was actually started at last year's meeting, when Henri Poole talked about meetup.com as a way to get like-minded people together, and later in the meeting somebody announced that they had just created a meetup group for fsf members. I signed up, and several months later the Ithaca meetup had accumulated enough members that it was possible to have meetings. It's been growing since then. (Mitch said that I started the group, which is not quite true). Knauth wrote "Ithaca Meetup" on the board. Through the rest of the day, people kept mentioning our meetup as an example of a successful community effort. We're famous now! (I made sure to write the meetup URL on the blackboard afterwards. The idea of meetup.com is that you sign up and indicate your location; as other people sign up, you'll see who and where they are, and the site helps you plan meetings and tell everybody about the time and place.

Henri Poole's talk was about meetups and their kin (his talk last year was on the same topic). He's very right that when people meet face-to-face, they see the existence and strenth of their community. My own example: running GNU/Linux and knowing that other people run it too, is unremarkable. But finding a LUG and eating pizza with a dozen local linux users really reinforces your interest in the subject, and in the community (making you more likely to start a LUG or a meetup or an IRC channel with other people you know or in other places you hang out).

Lunch featured some extremely tasty vegetarian sandwiches. Mmm lunch.

Gerald Sussman's talk was about "big ideas in engineering", namely interchangeable components with standardized interfaces and parameters. Components can themselves be made of components, thus simplifying and modularizing ways of drawing and talking about circuit designs. He emphasized that science and engineering are not really separable: scientists invent tools, and inventors discover science. A preoccupation with "intellectual property" disrupts the science-engineering feedback loop.

Most of this he conveyed in the first few minutes of the talk. The ridiculous/awesome part was the delivery of the rest of the talk as a whirlwind tour through advances in electrical engineering, from the discovery of static electricity up to the future where we'll be making circuits out of biological parts. He came into the room with a big bag full of transparencies, laid them out on the table, and for the next hour he slapped them on and off the projector. In the space of about three seconds, he would yank off his timeline, slap down a picture of the inventor, swap it for a diagram or three of the invention, and put the timeline back. I don't think he ever put down the wrong slide, dropped a slide, put two slides down at the same time, or fumbled with any of the slides in his hands. The visuals just happened. I am in more awe of this than Marty is about Lessig's slideshow. This is overhead projector mastery. A dying art.

I don't know how much of his visuals were planned in advance. It seems like he had a handful of slides he knew he would show, and a deskful of related images to throw up whenever and wherever he felt like interjecting some extra information. Can't do that with powerpoint.

Unlike your standard biological or corporate meeting, nobody yesterday used a standard powerpoint presentation. (In defense of biologists, they often have a lot of inherently visual data to present). Poole put up an html page or two, but aside from Sussman and Lessig (more on his later), all the speakers just talked unaccompanied, looking directly at the audience. (Weirdos!)

A heckler tried to claim that patents were never a good thing. (Moglen's comment: patents made more sense when the period of the monopoly was comparable to the amount of time it took a business to find the capital to start building a competing product). Neither Moglen nor Sussman agreed with this heckler, but he claimed that they wanted to believe it and that he could prove it was true. Sussman's response: "Yes, I would love it if it were true. But it's hard to prove things that are false." [applause]

I don't know how I can possibly summarize Eben Moglen's talk. When you listen to Moglen talk, you feel like he's telling you secrets: what Microsoft is doing next, what IBM is doing next, what Sun just did last week that nobody heard about, what happened in such-and-such a secret meeting between this CEO and that CEO. I expect that since he was saying it to >80 people and a video camera that it's all fair game to blog.

Last year Moglen's talk was mostly about the SCO case - SCO was trying to be threatening, had recently sent a subpoena to the FSF asking for everything anybody ever said to anybody, and was generally trying to make a nuisance of itself. In the meantime, SCO has kinda fizzled out, and while the subpoena cost the FSF a lot of money and time, it wasn't as bad as they had feared.

Over the past year, Moglen spent a lot of time countering SCO-generated FUD for various corporations; he thinks the FUD didn't have its intended effect, which was to cause boxen to be installed with Windows and Microsoft products instead of free software.

He noted that while IBM's goal was to win the case against SCO, which they are doing, the FSF's goal was a little different: to win against SCO in such a way that shows Microsoft it's not profitable to go around making more SCO's. He seemed confident that the FSF is succeeding in that. (Background: Microsoft probably bought this attack outright.)

Another important development in the past year is the Software Freedom Law Center that Moglen is starting and that bkuhn has defected to. Its mission is to provide legal services to nonprofit free software projects. Apache, Samba, and MySQL, to name a few, are important to many people and companies; they are "essential free software" although they aren't owned by the FSF. The SFLC will be able to help them.

The SFLC is getting ready to exist in a physical location. They are recruiting lawyers to join (specifically young geeky lawyers fresh out of law school) and are also recruiting clients whose potential legal problems are important ones. The SFLC will be mostly for advising, not so much for actual litigation (although they plan to keep some funds around in case somebody gets sued, as a stopgap until someone else can step in).

Microsoft has been engaging in a bunch of patent-related FUD, including "muggings" where some big-name Microsoft exec invites a company's CEO out to dinner, and the CEO is met by a team of lawyers who threaten to sue over some patents, only they won't tell you about it until you sign an NDA. The NDA forbids you from talking about anything with your lawyers, including the fact that there was an NDA, or else says that Microsoft gets to pick what lawyers you share it with. One CEO published his story anonymously in BuisinessWeek (I can't find a link, though).

Microsoft has also been dealing with Sun. The story goes something like this: at present, IBM has the only plant that can make sub-90-nanometer chips. They estimated that their only potential competitor is Fujitsu, who would need about $3 billion and two years to build a competing plant. Meanwhile, Sun cut a deal with Microsoft wherein Microsoft: MS gives Sun $1.6 billion and promises no patent infringement cases, and Sun will take $1 billion of that and purchase 1/3 of the capacity of Fujitsu's new plant. Of note is that the patent-infringement deal contained one exception: MS is still allowed to sue over StarOffice and OpenOffice.

On the subject of patents, he pointed out that IBM has been making some interesting kernel contributions lately - patches that claim to be making minor code edits but are possibly "inventing around patents".

Moglen mentioned a growing sentiment that patents are something that hurt your customers. IBM has proposed a patent commons where it's making many of its software patents available for royalty-free use. IBM doesn't want to totally give up its patent aggression since that would keep it from fighting back against patent aggression from other companies, but they like the idea of donating a few for public use. He made the analogy that IBM is on the right side of the war, but doesn't want the other guy holding all the grenades.

FIXME: Microsoft something about injunctions against distributing a product while somebody is something patent infringement. MS is scared that somebody will keep them from distributing Longhorn. The court rejected this, saying that Microsoft hasn't earned the right to comment on patents. [applause]

Microsoft's strategy involves patenting document formats, especially formats used by Office. Massachussetts didn't like this - they want public documents to be published in open file formats so that the records don't die when the software that produced them dies. Microsoft responded by [licensing the patent? FIXME] for read-only access.

Microsoft would rather see the European patent [thing FIXME] die rather than see it go through with "interoperability carve-outs".

At this point, I noticed that David Turner was playing a game of SET on his laptop. I must find this program.

We should expect to see patent aggression from Microsoft particularly in the UK and in New Zealand. Meanwhile, the FSF is not ready to participate in a patent war itself - at least in the sense that if such a thing happened in 2005, "we will not have chosen the time".

Microsoft, he said, is "one failed upgrade away from serious trouble". It's possible that Longhorn will be that upgrade. A 3-GHz chip is required to run its kernel. So, Microsoft has committed for a RTM (release to manufacturer) date of May 2006, so that machines with Longhorn installed can be sold for Christmas 2006. It's not clear that they will be able to make that deadline. According to plan, the consumer version is hitting the market before the server version, which may be a mistake. Longhorn's server version is supposed to be out in mid-2007, which will probably be more like early 2008. IBM isn't worried - they say they'll have everybody running free software on their servers by then. "Some of the largest and most powerful businesses" share our opinion of what software the majority of servers should be running in 2008.

Version 3 of the GPL is coming soon. Maybe 2006, "while the eye of Sauron is looking elsewhere". They're working on it now.

FIXME "there's a whole lot of friends coming" "we have earned some political capital and we're gonna spend it" "unfreedom" ${China + Wal-Mart} > ${Microsoft} and China is on our side

Sun is getting rid of their software business for now, essentially hoping the community will develop it while Sun focuses on high-end hardware. Then, they plan to steal back the software, which is under a not-really-free license. Developers aren't buying it, though: why would they develop for Solaris when they could put their effort into software that is actually free, that will belong to the community instead of to Sun? Meanwhile, companies that Sun hopes to sell hardware to are discovering that they can do better - orders of magnitude cheaper - by buying commodity PCs and running free software on them. The odds are not very good that these customers will switch back to expensive machines running expensive software.

FIXME Sun positioned themselves as a RedHat killer, which was a bad move, cause who in their right mind is afraid of RedHat?

RMS

.

Lessig

.

.

.

.

.

dinner

Marty is never, never, never going to let me forget that I dined with Larry Lessig and Eben Moglen at the Middle East that night. In our several IRC conversations since that time, he has consistently referred to them like "<still jealous>Eben Moglen</still jealous>". Actually I sat down at the table thinking "oh, there's Mitch from Ithaca, and bkuhn is at that table too", only to find when I sat down that the guy in the corner - whose back had been to me - was Lessig. (Moglen joined us later on). I'm not cool, just lucky!

help us

If all this sounds like fun and you want to attend next year's meeting, or if you just want to support the FSF and the things they do, please consider joining as an associate member (If you'd like to credit me as the person who inspired you to join, my member number is #1939). Most of the FSF's budget comes from associate member dues, so by joining you are really helping the FSF to do its job.