Since we're a bit late in the process, it'd be best if we can get proposal ideas posted as soon as possible (this week -- Monday 12th, Tuesday 13th preferably) -- in order to ensure a favorable review of our GSoC application.
Organization results will be posted on the 21st January.
I believe I can write pitches for project ideas: i.e. support testing fatalError (and friends) or improving Swift-testing to use closures instead of function. But estimating Project Size, Difficulty and Recommended skills is not something I feel comfortable with. Do you already need that info?
Cool, thanks for chiming in! Yes, anyone can propose ideas.
I'm a little bit concerned about those topics? Can you clarify if this does have an approach that was endorsed by the swift-testing maintainers? E.g. @grynspan
An important aspect of GSoC is to provide students a chance to apply work and land their changes in the actual upstream repositories. We'd be concerned if this was an experiment on the side that the swift-testing team was not comfortable with for example. The project should provide the students a tangible deliverable that they can later "brag" (in a positive sense) about that "That thing in Swift? I did that!", so I'm concerned about project ideas which may end up not getting into proper upstream.
Or to look at it from another angle, it would not be fair to the GSoC participant if their project is not actually going to be part of a "core" Swift project, as GSoC is supposed to be about teaching them to participate in an existing community and project.
// small caveat: some server project ideas don't contribute "to" a Swift project per se, but since the server ecosystem is all about this small mesh or packages, this is kind of the expected way to "be part of the swift server ecosystem".
I'm afraid so, we have to have complete project ideas -- including these three pieces of information. It is likely if someone is willing to mentor a project they should be able to judge these three aspects.
Yes, an idea should have a dedicated mentor that has agreed to make the time for it if a project would be selected.
We don't have much experience in mentors which are not full time working on Swift, but we'd like to give it a go -- if you're interested in it, please propose an idea. To actually run such project we'd have to get some of your personal data in order to enter it in the GSoC website etc. We could consider pairing up first-time mentors with mentors which already have done it in the past in order to make the first time easier.
@mhjacobson does not seem to have responded to anything since last September. It would be great to get Swift on FreeBSD. I would be happy to help in anyway I can, but I am not a compiler guy (these days, just a developer who uses Swift).