The thing is that making up a succesful proposal (with high chance to be accepted) requires outlining the schedule of what will be done. Maemo has a lot of project ideas, and competition will be high. So the proposal to improve shepherd should look as good as possible in the eyes of mentors. And as usual it should be done "by yesterday" so it's only several hours left to deadline (but that's totally my fault).

I couldn't find source code of shepherd anywhere, so I'll try to outline what tasks there are in the project and what progress is made. It's just my guesses, and sorry if I offend you with underestimation. :-) I'll rely on the following guesses in my proposal, but if you correct something, I can fix my proposal as well.

• A nice, non-geeky GUI in Qt - GUI currently is very basic, and the Qt features used suffer from platform bugs
• API for plugins - works, but needs improvement
• Plugins (both conditions and actions) - several basic ones are ready, creating more plugins is necessary
• Scheduling functionality - completed and very powerful
• Integration with the OS (in plugins) - needs improvement, so that user just clicks buttons and doesn't input scary numbers :-)
• integration with a mapping application for location-based conditions - not done

Did I miss or underestimated completion of something?