GSoC 2019 - Feedback request

Mojca Miklavec mojca at macports.org
Mon Apr 8 05:54:29 UTC 2019


Dear Amir,

On Sun, 7 Apr 2019 at 18:21, AmirAli Mashayekhi wrote:
>
> Dear Mojca (and everybody else in the mail list),
>
> Thank you for your response, Sorry for the misundrestanding of the proposal.
>
> As requested I also tried to expose myself to the whole workflow of Macports developement cycle.
> Here is a link for my pull request : https://github.com/macports/macports-ports/pull/4039

Thank you.
Just please change the commit message, force push to the branch and
the it's good to go (the Travis failures are unrelated and Azure is
happy, so that's fine).

> After spending the day to check the ports in the tree (this way I also found mlterm to update as I mentioned above), honestly I think I need a little bit of hint from the community to see and select which port right now fits my abilities. nevertheless I have updated my proposal's general structure, I hope it makes more sense for you.
> The only missing part atm is to fill the port name I will work on...

You may check https://trac.macports.org/wiki/Tickets and see which
ports have more bugs, update requests etc. filed. Esp. those that have
been waiting for years.

Candidates for GSOC could be:
- KDE
- QT5 (it works, but there is a lot of work to be done with respect to
older macOS version)
- MySQL (not enough for the summer, but it needs a maintainer), Pandoc, ...
- full ruby gems spectrum, full haskell / cabal spectrum, lisp, ...
- ... plenty of others, I just listed some from the top of my head.

Since you seem to be interested in virtualisation and security,
another option would be to figure out how to set up virtualisation for
Mac OS X 10.6-10.14 (ideally for all of them, but even if just some
are done, that's still a lot), use libvirt to start and stop them, and
then figure out how to integrate this into our buildbot setup to test
pull requests in a secure way. You would need to prove to be
standalone enough to perform the research and execution.

There's another proposal for other buildbot improvements this year,
but properly doing this one task would also take a while and could be
a separate task.

(You don't have much time left until the proposal deadline though.)

Mojca

> My proposal link : https://docs.google.com/document/d/1RUmFN7PSPqbGzH4h8Hqq-qkTFgS3XJ_bIrsq7GhikYE/edit?usp=sharing
>
> Best Regards,
> Amir
>
> On Apr 06, 2019, at 04:21 PM, Mojca Miklavec <mojca at macports.org> wrote:
>
> Dear Amir,
>
> Nice to see you at MacPorts!
>
> You should send your email to the macports development mailing list
> and you would get the answer there. (I'm forwarding it to the mailing
> list now, but you should subscribe for the future.)
>
> On Sat, 6 Apr 2019 at 20:14, AmirAli Mashayekhi wrote:
>
>
> Dear Macports GSoC admins,
>
>
> I have uploaded my proposal draft on GSoC webpage. Unfortunately we at EPITA(mon École) got informed late about this years GSoC opportunity. But this doesn’t affect my true excitement about participation, open source and contributing on them.
>
>
> Late arrival of proposals is not a problem per se, it just means that
> you have less time to improve your proposal and demonstrate your
> skills compared to students who started early, so you need to excel
> more to catch up.
>
> My friendly suggestion would be to not try to write three proposals in
> two days, as that will more likely decrease your chances of getting
> selected, rather than increase them. Not because students applying to
> other orgs would sound less serious, but because you hardly have
> sufficient time to write one good proposal, let alone three.
>
> I appreciate if there is any feedbacks so I can enhance it before finalising my application.
>
>
> You nicely answered the auxiliary questions (meant to help us
> understand the student better), but the project proposal is completely
> missing.
>
> The proposal fails to show any effort to research and understand the
> task. The description / plan is so vague that you could in theory
> submit a three-line patch to an arbitrary outdated port (for the whole
> summer) and claim that you have delivered what you promised. Not even
> a single piece of software was named as an example of what you would
> work with (after six years of using MacPorts you should at least be
> able to name one single piece of software that you wanted to use, but
> was too old or missing).
>
> In any case we are asking all of our candidate students to create some
> pull requests or small demo projects demonstrating their
> problem-solving skills. This would be the perfect opportunity to
> update & improve a couple of important outdated unmaintained ports of
> your interest.
>
> My draft should be available on your dashboard by now, but I shared a link here in case of comments.
>
> My draft link : https://docs.google.com/document/d/1RUmFN7PSPqbGzH4h8Hqq-qkTFgS3XJ_bIrsq7GhikYE/edit?usp=sharing
>
>
> Best Regards,
>
> Amir
>
>
> Please note that this was meant as a constructive feedback.
>
> Mojca


More information about the macports-dev mailing list