Re: Participating in Outreachy's December 2026 cohort

From: Christian Couder <christian.couder@gmail.com>
Date: Sep 5, 2026, 08:26 UTC
Message-ID: <CAP8UFD0oYnoXgQ84wHbGg3+QhX78Ucn_CXXYOe8uFpReb7X1Ng@mail.gmail.com>
In-Reply-To: <CAP8UFD367UD=AomNVHEBnhY-2DQmqTNRcBX6NW7YZywWgOmxTQ@mail.gmail.com>

On Wed, Aug 26, 2026 at 12:55 PM Christian Couder <christian.couder@gmail.com> wrote:

Show 6 quoted lines

> Open Source communities have until September 11, 2026, at 4 pm UTC to
> sign up to participate as mentoring organizations in Outreachy's
> December 2026 cohort (see
> https://www.outreachy.org/blog/2026-02-06/may-2026-call-for-mentoring-organizations/).
> And (co)mentors have until that same time to submit project
> descriptions.

Thank you Usman, Kaartic and Pablo for volunteering to co-mentor!

As no one objected to Git participating in Outreachy, I will submit a Community Application for Git very soon.

We will still have to sign up as mentors and find and submit projects for interns to work on soon. We have until September 11, 2026, at 4 pm UTC, which is less than one week from now.

I have been thinking about the following projects:

1. Continue removing global state (and perhaps start libifying some
parts of the code base).

I think Tian Yuchen said there is not much work left related to "environment.{c,h}". So maybe we need to adjust the project compared to its previous description or maybe make it about actually libifying some parts of the code base.

2. Improve how command arguments and options are scanned and parsed.

There are still a number of places in the code base where command arguments are scanned or parsed using ad hoc hand-rolled code instead of an existing API like the parse-options API or the early-scan API I recently proposed in https://lore.kernel.org/git/20260902161047.476753-1-christian.couder@gmail.com/. Those pieces of ad hoc parsing or scanning code are often buggy and difficult to maintain. I think porting them to an API while improving that API if necessary can be an interesting project and bring a lot of long term benefits to the project. As the parse-options API is already quite feature-full, there are likely some areas where porting the parsing code to it should be relatively easy which is newcomer friendly.

Let me know your opinion about these and if you have other project ideas.

> Last year the Git community mentored a single intern. For a long time
> GitHub or GitLab used to sponsor Outreachy interns working on Git, but
> last year both declined to do it, so the intern was sponsored by the
> Git project itself.

I will start poking people at GitLab about this, but it would be nice if we could get some kind of help with this regarding other companies too.

Thanks, Christian.