Some of these probably aren't really that suitable:
After this filtering, we seem to be left with:
- An easy way to share wiki content on social media services
I don't think this is a good idea unless the political climate on wikipedia has changed (we want the gsocer to stay. Flame wars are probably not conducive to that)
- Write an extension to support XML Sitemaps without using command line
Meh, maybe. Rather low impact. Very few people fall into the use case, its debatable how much a sitemap would even help with seo. Its also not the easiest thing to do to make mw work on a schedule, especially if the task needs to be accomplished a little at a time.
- Extension:OEmbedProvider
Would be concerned that its not likely for the student to see their work used (deployed) after a single summer of work unless it was mentored by someone who would have the ability to deploy the extension themselves. That's obviously not a requirement and someone could do a lot of good work in the direction of making it useful, but something to keep in mind.
- Allow smoother and easier Wikimedia Commons pictures discovery
This is more - come up with cool ideas (which is needed) rather then an implementation project. Is that ok for a gsoc project?
-bawolff