gsoc: Improve OpenDocument in KWord

Just like last year the KDE-family did provide us again a gsoc-slot to work on Improve ISO OpenDocument support (in KOffice, particularly in KWord). Not only did the number of mentors drastically increased but personally I also had the feeling the number of very good proposals did. To be allowed to read all of them was such exciting that I found myself spending hours over hours on the huge list reading lots of proposals again and again just for fun. To be able to see such a pool of impressing ideas and developers was already reason enough for me to participate again as mentor rather then as student. Thanks to google for this unique opportunity!

One of the things I totally dislike about the mentor-role is the need to coordinate with all the other mentors on the try to "rank" those proposals somehow what is needed since we can't just select all of them or at least those who are very good but we are limited to a more or less fixed number. The fate of having so much good choices. I also remember my own frustration in 2005 and 2006 where none of my proposals got selected (though looking with my current experience at them they didn't matched into the "very good" category but at the best case in the "good" one). I guess part of that dislike is also the need to base such a decision only on the proposal itself which is text- rather then code-based and therefore, in my eyes, much more difficult to parse. One thing what helps here a lot is to see some kind of planing the student made. So, dates with milestones that allows to see what to expect next few months and probably even some details how the job should be technical achieved - but that may only my biased view on things. Another thing I dislike is that gsoc happens during the summertime which is not exactly the best time to be productive cause of all those sun out there. A gwoc (winter of code) would maybe match better :)

I guess it's quit usual that everything has two sides and once those ugly selection-burden got taken, the real work and with it the fun starts/continues. As mentor it's important to guide the student specially at the beginning. It may needed to help him to setup a developing-environment (up-to-date trunk, SVN-account, blog, etc.), to guide through the probably already existing code-base, to coordinate (align times, provide a path to start to walk on, offer co-mentors that can help if the mentor himself turns ill, has a hang-over or just needs his weekly sleep-phase to reboot), etc. Really a lot of initial work but it pays out if you see the commits rolling in. That's not only a thing of code but also of a great feeling to see that at least some of the experience collected within last years went "upstream" (probably a similar kind of feeling school-teachers have and what may drive them to continue there underpayed job?). Over the time it may turn from teaching over pair-programming like teamwork till being teached what is then another very great experience.

Back to the topic this blog started with; So, I'll mentor the improving OpenDocument-support and will try this time to blog a bit more about the progress we are doing there since there are rumours out that not everybody does monitor those thousands of commits KDE has each week ;)

After spending the last few days to discuss and plan what needs to be done to solve that rather complex task the most effective way, we did land a first series of commits for this years gsoc just yesterday and that even 2 weeks before the official gsoc-start. Thanks goes here to Pinaraf, the student that does mentor me (or was it the other way around? guess not since he did participate already at gsoc and KDE before and we where able to skip some of the inital steps and are atm already in the teamwork-mode), the KOffice-developers for providing again such a great environment to work in, the KDE-family for there trust in us and the believe in open and free documents and last but not least google for making all this possible. You all rock!