GSoC libgit2 bindings

Saurabh Kumar saurabhgeek92 at gmail.com
Fri Apr 26 16:54:45 UTC 2013


Jonathan Duke Leto writes:

> Howdy Saurabh,
>
> [..]
>
> Good question! We can make this this as easy or hard as we need, to make it
> doable in the timeframe we have and interesting to you.
>
> Only do the following things:
>
> 0) report the Rosella bug
> 1) build parrot-libgit2 as far as you can, until you hit the problem with
> Rosella. You should still be able to install and use parrot-libgit2 without
> Rosella.
> 2) get familiar with Winxed and PIR. parrot-libgit2
> 3) See if the unify_build_system branch works for you:
> https://github.com/letolabs/parrot-libgit2/compare/master...unify_build_system(ask
> us if you need help testing a branch properly)
> 4) Specifically don't go into libgit2 madness until you have your version
> control and build tools mastered.
>
> Part of your proposal should be about updating parrot-libgit2 to the latest
> stable release of libgit2. This will involve changing some of the API
> binding function signatures, documentation, tests and examples.
>
> Please start by making a detailed proposal that has one paragraph for each
> week of the program. More detail is better than not enough.
>
> Let me know if you have any more questions!
>
> Duke

I've already discussed some of these things on IRC(my nick is sa1).
Since then I have read a little documentation on winxed, NCI and libgit2
and while it has not been enough I managed to write a first draft of a
proposal[1], which mostly consists of the schedule at the moment.

It's currently a little naive and lacking details and I'm not sure if
I've spread out my work as I should. I will continue changing the scope
and adding details in further iterations.

I've spent the last day focusing on winxed and NCI and not so much on
libgit2. I only have vague doubts(and a vague understanding of PIR) for
the moment. Would additional work be required for writing OO interfaces in
Perl6? Since PMC would would already be present, it seems to me that it
should be trivial.

In addition to general feedback about my proposal, I would also like
some specific feedback on what you think is the heaviest part of my
work, and how much time it would take in contrast to what I have already
alloted to it.

[1]: https://gist.github.com/sa1/5468408
-- 
Saurabh Kumar


More information about the parrot-dev mailing list