TedPage: In and out of Sodipodi

Well, peer pressure works, so gather round kids for story time. I'm going to take you back long long ago...

After graduating from college I was burnt out, really burnt out. Basically I did nothing for at least a year, probably longer. I can't remember what I did really, but I can't find any evidence of something productive from that time. After being forced to take a week off of work, and having nothing to do, I realized that I needed a hobby. I wanted to help out with an open source project, but I didn't know how or where, so I started looking.

First project that I did anything for was Tenes Empanadas Graciela, which is a game similar to Risk for Linux. I didn't help out that much, mostly a just made RPM packages for PowerPC. I really like the game and still play it today. Though, the AI is getting better faster than I am.

I had, at one point, thought about writing a vector illustration program. I had a design, and a logo, and even if I knew where all that was I wouldn't show it to you. I decided to try and help the Sodipodi project by reading the mailing list and finding a bug to fix. Here is my first Sodipodi patch. This happened in the middle of a summer vacation from computers that Lauris was taking. I started out frustrated with the progress of Sodipodi. I plugged along, and started working on a plug-in system for Sodipodi. When Lauris wasn't getting around to including my patches, I figured it would be better to document them, and that started the Doxygen documentation that we have today on Inkscape. Much of my plug-in work did eventually get included in Sodipodi, but not without long delays and a large amount of effort.

I was getting frustrated with Sodipodi, and I was still looking for a project to call home. I'd also started lurking on the GStreamer mailing lists as it was similar to parts of my senior project (I still lurk there today). Iain announced that he was looking for a new GNOME-media maintainer; I volunteered and became the GNOME-media maintainer.

I was a horrible maintainer. Well, maybe not horrible. The package didn't digress, but it didn't progress the way it should have either. There were several patches that people had written and put into Bugzilla that I never got to in my tenure. This was for a variety of reasons including family issues, an underpowered laptop and in the end because Inkscape was being born. Eventually, I found new maintainers and gave up GNOME-media.

The reason that this back story is important is based on how GNOME CVS works. When you get an account for GNOME CVS, you can edit any project on the server even though it is bad form to edit any where the maintainer hadn't given you permission. So, when I was GNOME-media maintainer I got a CVS account, which allowed me to modify Sodipodi's repository also. I never committed any of my patches to the mainline branch, but it came in useful as the hydra branch got started.

I think it is important to reiterate why the hydra branch was created. The goal was to provide a staging area for patches to Sodipodi, this way Lauris could approve them as a bunch, hopefully making it easier on him. The model was similar to the -ac and -mm branches on the Linux kernel. Since I was the one with CVS access, I became the bottleneck integrator of the patches. There were quite a few, but at this time I was much more familiar with reading diff's and using autotools. I can't say that I read them all though, and I understood very little of the Sodipodi codebase, that's what teamwork is for.

As the hydra branch matured we wanted to see the code get used, we even considered releasing it as a separate branch. Lauris didn't seem to be interested in the code or the branch. And, well, we started Inkscape from that branch. To his credit, Lauris did integrate most of the changes in the hydra branch into the Sodipodi mainline and made a release with those changes after the fork. But, at that time it was too late; Inkscape was born.

When it was decided to start Inkscape, things started going really fast, and we never looked back. One of my favorite memories from that time was when Bryce said that he'd consider the project very successful if we had 100 downloads a day. Bryce, we're very successful. But, that can't be attributed to just those who started Inkscape, more developers, testers and users come on board everyday to make Inkscape better. I may be attributed as a founder, but Inkscape isn't about individuals.