Talk:Outreach programs/Selection process

Not sure where to comment given this is on a talk page already, so I'll put this here. I was really glad to see "Is there proof of previous contributions or interest in Wikimedia projects". This is purely anecdotal, but it seems to me that people who are active editors in some wikimedia project, and join some mentorship program in order to work towards fixing an issue that directly affects them (and their community) tend to stick for the long term, and care more about their project. Bawolff (talk) 22:59, 20 November 2012 (UTC)

I'm not comfortable with the three last criterias. They are too formal and too much criteria will lead to a restrictive choice, whereas everywhere at Wikimedia, we always tried to get a diversified and rich set of people.

Outreach programs are also made to gather new, external contributors. But it's offered to add "Is there proof of previous contributions or interest in Wikimedia projects?" criteria. This seems incoherent.

The next criteria is complicated to evaluate beforehand. New works lead to new opportunities, sometimes randomly, and we don't have a crystal ball to determine what will occur.

The last criteria isn't coherent with empiric community observations. There is a regular flow of arrivals and departures in online communities. The departure decision could be an exit following an issue, a decrease of motivations or other priorities in life taken over project participation. So, I don't really know how you would like to evaluate this criteria for a newcomer. A not really motivated or shy newcomer could become an hardcore contributor like a very motivated person could leave quickly after. --Dereckson (talk) 22:31, 21 November 2012 (UTC)

These criteria help us evaluating a candidate and a proposal as a whole. We are not trying to build any scorecard to be measured with a calculator. Knowing about previous interest and participation in Wikimedia projects does help getting a better picture about candidates and the familiarity they might have with Wikipedia et al as users. 9 and 10 would be based on what the own candidate thinks and says, not on ourselves trying to secretly guess it. It is interesting to know whether someone sees a mentoring program fitting in their career path and personal interest or they are simply trying something anew just because.

There is no right or wrong answers in these criteria, but they help seeing the consistency and potential risk of a specific candidate / proposal. One person might represent a wild bet, yet original and convincing. Other candidates might look like a very safe bet, but still be beaten by a potentially risky candidate. It will be up to the mentors to go for one person or another, based on their perceptions on the sum of all these criteria. I believe these questions are useful for the candidates themselves preparing their projects.--Qgil (talk) 19:13, 26 November 2012 (UTC)

Criteria 2 and 6 make sense for coding tasks but feel not totally applicable for documentation/marketing/bugtriaging/etc. tasks. It's required that there should be "finishable" subtasks for ever-ongoing activities like improving documentation or bug triage, but I have problems to come up with quantifiable criteria for bug triage, as we're after quality in the end and not "comment on X tasks in 3 months". Ideas welcome. --Malyacko (talk) 18:29, 27 November 2012 (UTC)

Now they are 3 and 7. I think the criteria still apply (flexibly) for non-coding tasks. As long as they require a clear deliverable at the end of the program, which I think is a requirement anyway for any type of project. I'm not a big believer of ongoing tasks. "Triaging bugs" feels vague anyway. If you still want to push in that direction you could go for something like "Go through all the open MediaWiki Core bug reports older than MM-YYYY".--Qgil (talk) 16:31, 28 November 2012 (UTC)

I have to say I'm somewhat in agreement with Dereckson's comment that the last three criteria don't fit, though for a different reason. Evaluating the motives and long-term goals with respect to the candidate and the project seems like it's not in keeping with the usual free software development model. People aren't supposed to have long-term interest in a project when they start hacking on something--it's usually just an instance of jumping in to build a feature, or fix a bug, and then jumping back to whatever else you were doing. Encouraging that mentality should definitely be part of most free software mentoring programs. The long-term benefit to WMF will likely be more in the terms of the features that get added, and the large number of newcomers that we guide through the orientation process. That many new developers will wind up helping to smooth rough edges. --216.38.130.166 17:59, 28 November 2012 (UTC)

Alright, I have softened 9 simply asking for any previous Wikimedia contribution of any kind (always good to know). And I have removed any reference to long term interests of commitments, still asking why is the candidate interested in joining the program here and now (good to know in any case). The feedback was useful. I hope with this we are all happy about the list of criteria now.--Qgil (talk) 21:56, 28 November 2012 (UTC)

What does "Are there 1+1 mentors committed to support the candidate through the program, and beyond?" mean? Specifically, does "1+1" mean that each candidate has 2 mentors? I'd support that. Sharihareswara (WMF) (talk) 01:57, 28 November 2012 (UTC)

That was my cryptic way to express "Ideally one primary mentor and a secondary". I wish we could simply write "2 mentors" but I fear many proposals being turned down even having one great mentor available just because we don't have a second. I'll think of a better rephrasing keeping still one sentence short.--Qgil (talk) 16:34, 28 November 2012 (UTC)

• Let's make sure that there's a reasonable expectation, for mentors, participants, and the org admin, re: whether we have an application discussion period. For example, if the mentors are looking at the applications and realize they'd like to have a phone interview with the applicant, ask her for her past open source commits, etc., etc., it would be nice to have a time period marked out for that kind of iteration and information gathering.

I have rephrased a bit the Process section, trying to be more explicit in the points you mention.--Qgil (talk) 16:48, 28 November 2012 (UTC)

• Also, we need to ensure that the project plan part of students' applications are someplace public and world-readable before the application deadline, so that other mentors and org admins can check them and give feedback. Sharihareswara (WMF) (talk) 01:56, 28 November 2012 (UTC)

Good point to put it explicitly instead of assuming. I have rephrased the second point: "Is the proposal publicly available and has it registered community interest?"--Qgil (talk) 16:39, 28 November 2012 (UTC)

Melange for Outreachy?

There was a question on IRC regarding this page: "should I have an account on melange for outreachy?" " Well, it was written here https://www.mediawiki.org/wiki/Outreach_programs/Selection_process#FAQs_for_candidates , which created confusion. I guess it is only for GSOC. Thanks". Either this part should be made GSoC-specific or clarified.  « Saper // talk »  07:47, 29 October 2015 (UTC)

The equivalent of Melange for Outreachy is https://outreachy.gnome.org/ . When I was a mentor, I had to confirm something there but it was never really used. Nemo 15:18, 29 October 2015 (UTC)