A bunch of us Inkscapers are setting up blogs and rss feeds and such, so I figured I'd be lazy and just dust off my advogato account. ;-)

Just to catch my diary up to date... I decided my fate in life was to become an Open Source developer and help organize Open Source projects like WorldForge. So, I moved up to Portland and as my contract with Morgan Stanley ended I switched over to the newly formed "Open Source Development Labs" (OSDL) in spring 2001. I focused first on helping assemble the company's infrstructure, databases, services, etc. but last year moved into a Performance Engineering role, where I help work to improve open source software through testing.

Meanwhile, I got burnt out of doing coordination for WorldForge and got involved with Gnupedia/Nupedia, which were aimed at creating an online, Open Source style encyclopedia, but weren't quite working. Several of us then started trying doing this using wiki software, and *that* worked well. Thus Wikipedia was born. :-)

But I got tired of arguing with folks on Wikipedia, and wanted to get back into something coding-oriented anyway, and so joined another project called Sodipodi. After a while it looked like we could really turn it into an open community-oriented project like WorldForge had been, but that approach didn't really jibe with the Sodipodi structure, so instead several of us struck out on our own and formed a new project called Inkscape.

Since I absolutely cannot draw, I helped form another project called the Open Clip Art Library, to help skilled artists help the rest of us. We have a monthly release schedule, and are included in several Linux distros.

At work, my most recent project is organizing an effort to test the Linux NFSv4 kernel code. NFSv4 marks a major change from NFSv3 in that it adds 'state' tracking, security, and integrated caching. NFSv3 has always been infamous for its insecurity; it's lack of state awareness has also led to some inherent performance limitations that had been preventing its use for a number of important applications. Thus, there's some major benefits to NFSv4 that people are looking forward to.

However, the challenge is that switching to NFSv4 is going to be a pretty intensive change for end-user companies; they have to weigh the benefits against the risk of bugs, downtime, and the normal frustrations of new code. Thus, the name of the game here is Testing. If we can test the heck out of NFSv4, we can both eliminate problems and give users quantification of its stability and benefits. We want to prove that it's safe to migrate to, and if it isn't safe, we want to fix it so that it is.

The way I'm going to try to achieve this is to build up an open source style community engaged in testing of it, like I've done before. The thing that will be interesting about this is that it's much more corporate focused. However, my intuition says that the same techniques for building an open source project for games/drawing applications/encyclopedias can be adapted for building one around network file systems. So far so good; we've got half a dozen companies committed to working on it, and we're well organized to include more.