It was a beautiful beginning of Summer. The Google Summer of Code was on and it was glorious and fun, or at least everyone expected it to be after our previous two years in the program went so well. There was no reason to worry. We had all our ducks lined up in a row – four slots assigned by Google to our team, all our infrastructure set up just right, and friendly and competent mentors paired up with promising and enthusiastic students who were going to work on great projects. However, “The best laid schemes of mice and men Go often askew.”

On the bright side, Thach Tran has completed his project successfully, which put Scribus’s PDF export once more ahead of most closed-source DTP programs. I would like to thank him for sticking around, working through the harder parts of the project, and persevering to the end! His mentor Pierre Marchand commended Thach on doing a great job of keeping his branch in sync with the trunk. He also kept lines of communication open and made frequent appearances in our IRC channel. All of this made Pierre’s job easier as the new project code could be compiled at virtually any time and any problems could be communicated quickly. So, congratulations Thach Tran.

Having said that, the rest of our GSoC 2009 student team has not fared very well. The second student did make an effort to move his project forward. Unfortunately, it seems that the project was outside of the scope of his skills at the time, so we had to fail him at mid-term evaluation time. We wish him good luck in his future endeavors. This is possibly the most frustrating part of the GSOC 2009 experience for us as it shows that our process of vetting potential student coders was not robust enough. If we get to do this again in the future I believe that we won’t accept any students who have not produced any patches to the Scribus codebase. There is always a churn of low-hanging fruit in the code as new features introduce easy-to-fix bugs and our fantastic user community constantly comes up with feature requests and bug reports. So, it wouldn’t be a problem for a competent student to produce a patch or two. I think it would save us a tremendous amount of time and effort and reduce Google’s overhead. Even if we only get one candidate with patches I’d rather relinquish the remaining slots to other organizations than risk accepting people who just won’t be able to make it even with hand holding by mentors.

The remaining two students bailed out before getting anything done at all. I suppose there could’ve been some objective reasons for their early departure, but they have not come up with any, so my conclusion is that they simply wasted our time. Unfortunately, they made it ever so difficult for other people with similar backgrounds to ever get enough credibility for acceptance into the program. Later on, one of the students made an appearance and promised to finish the project outside of GSoC. We’ll see if that promise turns into something tangible.

The story wouldn’t be complete without a mention of the Google folks. As before, they ran a fantastic program and provided all the help and resources to the GSoC organizations needed for success. I cannot thank Leslie Hawthorn, Ellen Ko, Cat Allman and the rest of the Google SOC crew enough for giving so much of their life to this program. Thank you guys!!! I’d also like to mention one of Melange gurus Pawel Solyga who was very prompt in responding to a bug report. All in all, the world is a better place thanks to the Google Open Source team.



Scribus Team GSoC 2009 Administrator