List of things that should stay.
- Project flexibility. It's really nice you can change the to-do's for your project, to not get stuck with certain problems, and get an optimal result. I completely ditched the original idea of the Semantic Layers extension after some discussion with my mentor, and went for Semantic Maps instead. Half way through, it even got slip into 2 parts. --Jeroen De Dauw 00:12, 20 August 2009 (UTC)
- The awesome project idea's. This made it really easy to create an initial project plan for me. --Jeroen De Dauw 00:12, 20 August 2009 (UTC)
List of things that could be improved upon for next years GSoC.
- Community support. Wikimedia has a lot of enthusiastic people that have no problem helping a GSoC student out. A central resource for the students is lacking though. It would be awesome if some sort of MediaWiki GSoC group was created at the start of the program, containing all students and all mentors. This would have the following advantages:
- Students can share idea's with each other, and share solutions for common 'new dev' problems.
- Students can get support from people who are familiar with the MediaWiki code, different from their mentor. It's better to have multiple strong contacts with the community, and have more then one person review your plans or code. You also don't get stuck when your mentor becomes temporary unavailable by going on holiday or something similar.
- If a student has problems with his mentor, he doesn't get cut off.
- The more contacts a student has, and the more he gets support, the more likely he is to remain active in the community afterwards.
- Such a group could remain after GSoC ends, and would then form a great resource for new students (and mentors) the following year. It could be used to make the whole process more visible and transparent, and so draw in more mentors and students.
- It's fun!
That's my opinion on it anyway --Jeroen De Dauw 00:12, 20 August 2009 (UTC)
One thing I've thought about recently is the possiblity of Wikimedia Foundation sponsoring students independently. This year we were lucky to have basically the same number of GSoC slots, interested mentors and serious proposals from students: four each. In subsequent years that could change, and there could be more demand than available slots. It would be great if Wikimedia were able to fund such "overflow" projects on itsown - the amount of money needed is fairly small for an organization like Wikimedia, and the potential reward is quite high. Besides the money, Google also provides the structure, of course, but I don't think that would be hard to replicate locally, on an informal basis - there's no need for a system that tracks progress, relays communications, etc.; it could all be done via wiki pages and email. Yaron Koren 15:12, 20 August 2009 (UTC)