Re: What's cooking in git.git (Aug 2025, #05; Mon, 11)

From: Taylor Blau <me@ttaylorr.com>
Date: Aug 12, 2025, 14:41 UTC
Message-ID: <aJtSqgJ7w02Ox74w@nand.local>
In-Reply-To: <aJs5Gee3ZVCJX8dk@pks.im>

On Tue, Aug 12, 2025 at 02:52:41PM +0200, Patrick Steinhardt wrote:

Show 20 quoted lines

> > * ps/commit-graph-wo-globals (2025-08-07) 10 commits
> >  - commit-graph: stop passing in redundant repository
> >  - commit-graph: stop using `the_repository`
> >  - commit-graph: stop using `the_hash_algo`
> >  - commit-graph: refactor `parse_commit_graph()` to take a repository
> >  - commit-graph: store the hash algorithm instead of its length
> >  - commit-graph: stop using `the_hash_algo` via macros
> >  - commit-graph: fix sign comparison warnings
> >  - commit-graph: fix type for some write options
> >  - commit-graph: stop using signed integers to count Bloom filters
> >  - trace2: introduce function to trace unsigned integers
> >
> >  Remove dependency on the_repository and other globals from the
> >  commit-graph code, and other changes unrelated to de-globaling.
> >
> >  Will merge to 'next'?
> >  source: <20250807-b4-pks-commit-graph-wo-the-repository-v3-0-82edef830a1e@pks.im>
>
> I don't intend to reroll this series for now. As long as you are happy
> with the signedness-related patches I think this should be ready.

I am still not sold on the first four of these patches, and I share Junio's concern[1] that the "int -> unsigned int" changes are not well justified.

As a practical concern, the "max_commits" and "size_mult" values should never come even close to INT_MAX, so I am not sure that the wider range is giving us all that much. I am a little more convinced by the Bloom filter changes, but since they are purely for debugging and also exceedingly unlikely to exceed the signed INT_MAX, I do not think they are absolutely necessary.

That said, I don't feel strongly enough about the lack of justification here to hold up this series[^2], so I am fine with it moving forward if both you and Junio are happy with it as-is. But I am left wanting a stronger justification for the first half of the changes.

Thanks, Taylor

[1]: https://lore.kernel.org/git/xmqqpld6kb4t.fsf@gitster.g

[^2]: I don't want to waste the list's time debating a signed-ness conversion when we have much bigger fish to fry, but I also do not want to deviate too far from our usual standards, either.