Looking for issues and contribution

[Kshitij Pandey]

Hi,

I'm Kshitij and I'm very interested in contributing to git.
I'm a bit overwhelmed and do not know what issue(s) to start with. My primary operating system is Windows and secondry mint os and I have knowledge in Python, JavaScript, C++,  Git, Java, nodjs,mongoDB and flutter(learning recently)

I would really love some help with any recommendations of good newcomer issues that relate to my current skills as well as resources I should checkout as a newcomer (in case I missed any).

Cheers,
Kshitij

[Konstantin Khomoutov]

Hi!

Not to discourage you, but really, given the fact you did not manage to find
contributor's guidelines for Git, Git for Windows (and Chromium you have
referred to in another message to this list, which has nothing to do with that
project), I would say you're not yet on par to sensibly consribute there
because these projects are reasonably hard-core (in different respects, but
this does not matter).

I do understand that you're probably concerned with creating a portfolio which
would supposedly ease job applications for you, and that is okay, but please
consider starting with lower profile F/OSS projects before embarking on top
tier ones. I mean, when you will be ready to embark on such projects, you
won't be asking questions like this one I'm answering ;-)


To make things clear, everything written above is my personal opinion, and,
when it comes to Git, Git for Windows (and Chrome) I'm a mere bystander.
I happen to contribute to Git by trying to help people out on the mailing
lists related to it, but that is all. I mean, the actual developers of Git and
Git for Windows can of course maintain an opinion different from mine.
Still, my opinion is based on being involved in F/OSS for, like, more than 20
years, so I think I do have certain perspective.

[Philip Oakley]

Hi Kshitij,

Git is a bit of a beast with some soft edges. By that I mean that Git provides a part of the critical infrastructure of Linux and many other software projects by securely storing the version history of those projects, so great care is extended to keeping that promise.

The project doesn't present itself as well as some may hope, partly because it's by developers for developers. It's not trying to sell itself. Konstantin has given some pointers to finding some of the start points in the codebase.

If you want to have a look at the project, then do look at its mailing list archive at https://lore.kernel.org/git/, or register to receive mails at http://vger.kernel.org/vger-lists.html#git (i.e. request 'majordomo' to send you emails), to see how patches are submitted and responded to.

You can search the archives for '#leftoverbits' or 'GSOC' (Google Summer Of Code) (and other outreachy activities).

One area I recently looked at, if you are interested in a niche project, was the 'pretty-formats' column alignment for wide characters, which is currently faulty. see https://lore.kernel.org/git/20230119181827.13...@iee.email/

My analysis suggested that the code needs an extra step that counts, within the string that might be displayed, display columns as well as the regular 'char' count, resulting in an array of pointers to the displayable 'glyphs' (including their composable parts), the zero width terminal display codes, and the final NULL terminator, plus a matching array of column widths of those display glyphs.

If you do look at the code, note that it recently had a security fix for a potential 'int' vs 'size_t' buffer over-run error for systems that use 32-bit ints of a 64-bit pointer machine.

Philip