Hi, long email but only for GSoC mentors and curious minds alike.

WARNING: GSoC & common sense requires absolute confidentiality about discussions or resolutions of candidates. You can't share any confidential information, no matter how evident it looks to you or how well you get along with a student or anybody without mentor / org admin access to Wikimedia GSoC. Google forbids explicitly any leakage of information before they publish officially the results on May 27.

We can and we must discuss publicly our selection process, though.

Ideally GSoC mentors would have a private call and discuss until deciding on a ranking of candidates. But with 38 people in different timezones, 47 proposals and ? slots this clearly won't work.

Some organizations resolve this situation with votes, but I don't think this is a good solution in our context and the Wikimedia community favors consensus over voting anyway.

Your distributed feedback on essential/desirable features has been very useful to make a first decision. Let's try a second round of distributed feedback to solve the clear cases:

If you would be the only one deciding, how would you rank the proposals received (see the list below)? Please send me a PRIVATE email (not to this list!) with your ranking of features, ideally before the end of tomorrow Tuesday.

* Read https://www.mediawiki.org/wiki/Mentorship_programs/Selection_process (just updated) and act accordingly.

* Rank at least the projects you would prioritize before the one(s) you want to mentor. All the better if you rank more. Don't rank based on the title alone. All proposals have mentors feedback by now.

* No need to decide on specific candidates for the features you are not mentoring. It is enough to rank "Feature X", without you having to decide which of the students proposing Feature X should be selected.

* But you do need to specify which student you select for the 1-2 projects you are co-mentoring. Agree the names with your co-mentors. You can't be in more than 2 projects, and ideally in just one.

Mentors are free to skip the ranking game and go directly for the call. In that case their proposals will be ranked based on the feedback from the rest of us.

I will consolidate sensibly all this feedback on Wednesday, in a private document shared with the mentors. Hopefully some proposals will be clear candidates to be accepted or declined. Then we will also know how many slots we are getting from Google, and we can focus the discussion in one call or more with the mentors of the unclear cases.

Should work. We'll see.

The list of projects to rank:

* Android app for MediaWiki translation * Auto suggestion of categories * Automatic category redirects * Bayesan Spam Filter * Centralized Search Engine * Contribute to Wikimedia * Curriculum Wiki * Entity Suggester for Wikidata * Improve support for book structures * Improvement of glossary tools * Incremental data dumps * Incremental updates for Kiwix * Internationalization and Right-To-Left Support in VisualEditor * jQuery.IME extensions for Firefox and Chrome * jQuery.IME next big release improvements * Language Coverage Matrix Dashboard * MediaWiki API 2.0 * MediaWiki-Moodle extension * Mobilize Wikidata * Pronunciation Recording Extension * Prototyping inline comments * Refactoring of Proofread Page extension * Section handling in Semantic Forms * UploadWizard: Book upload customization * VisualEditor Math Equation Plugin * VisualEditor plugin for source code editing * VisualEditor plugins * Wikidata features * Wikidata language fallback and conversion * Wikipedia - My Encyclopedia