Re: What's cooking in git.git (Oct 2026, #01)

[Patrick Steinhardt <ps@pks.im>]

On Thu, Oct 01, 2026 at 03:48:09PM -0700, Junio C Hamano wrote:
> * ps/reftable-reflog-timezone (2026-09-30) 3 commits
>  - refs/reftable: fix on-disk representation of reflog timezones
>  - t/helper: fix segfault in "dump-reftable -t"
>  - date: add helpers to convert between "+HHMM" timezones and minutes
> 
>  The internal representation of timezones in the reflog for the
>  reftable format has been fixed to match the specification, which
>  dictates an offset in minutes rather than the parsed "+HHMM" integer
>  representation.
> 
>  Will merge to 'next'?
>  cf. <CAOLa=ZRVt=e3MqjLY=UkitfSg_YjpsfFjeDsEmqJQm-5YhopxA@mail.gmail.com>
>  cf. <xmqq33up73dm.fsf@gitster.g>
>  source: <20261001-pks-reftables-fix-timezone-format-v2-0-a4fd1f7cd21a@pks.im>

Yeah, this one should be ready.

> * mc/refs-rename-transaction (2026-09-23) 1 commit
>  - refs: run copy and rename through transactions
> 
>  Both reference rename and copy operations have been refactored to
>  use the generic reference transaction API, ensuring that hooks
>  observe the updates as unified logical operations rather than
>  piece-meal internal deletions and additions.
> 
>  Needs review.
>  source: <20260923133651.74120-1-maciej.ciemborowicz@gmail.com>

Oh, I completely missed this series. I'll have a look.

> * td/ci-large-test-resources (2026-09-30) 2 commits
>  - ci: use twice the CPU count on both providers
>  - t4205: compare huge output without diff
> 
>  CI resource exhaustion during test runs on GitHub Actions has been
>  mitigated by switching a comparison of a huge output to use a binary
>  comparison, and by capping the parallel jobs on Linux to the number
>  of available CPUs.
> 
>  Will merge to 'next'?
>  cf. <ar0gon2VE0RlG_cC@pks.im>
>  source: <20260930-ci-large-test-resources-v3-0-d65ac7c21b5f@gmail.com>

I'm happy with this series.

> * mc/refs-hook-report-old-values (2026-09-24) 1 commit
>  - refs: report old values to transaction hooks
> 
>  The 'reference-transaction' hook has been updated to report the
>  observed old object IDs and symref targets for unconditional ref
>  updates and deletions, instead of reporting the null object ID.
> 
>  Needs review.
>  source: <2af3eeadd18806c5298d53072428656885cff89d.1790269745.git.maciej.ciemborowicz@gmail.com>

Yeah, this is in my backlog to review.

> * ps/meson-improvements (2026-09-24) 7 commits
>  - gitlab-ci: fix hanging MSVC jobs
>  - meson: update wrappers
>  - meson: fix outdated completion helpers
>  - meson: use precompiled headers for unit tests
>  - meson: use precompiled headers for our test-helper
>  - meson: don't recompile git-remote-http(1) multiple times for tests
>  - meson: avoid recompiling HTTP sources several times
> 
>  The build configurations for Meson have been optimized to avoid
>  recompiling HTTP sources multiple times and to utilize precompiled
>  headers for test-helpers and unit tests, reducing clean build times.
>  Additionally, the shell completion tests and GitLab CI MSVC runner
>  jobs have been fixed.
> 
>  Will merge to 'next'?
>  cf. <CAOLa=ZS8Exa_WkMLiFNspeYaKFf40eOM19ZgecC=yTnFEAhMzQ@mail.gmail.com>
>  source: <20260924-pks-meson-improvements-v1-0-90b7f79f1c4e@pks.im>

Yup, I'd be happy to have this merged so that our pipelines are finally
green again :)

> * ps/setup-enforce-repo-passed-to-create-repository-has-no-state (2026-09-28) 7 commits
>  - setup: enforce that passed-in repo does not carry relevant state
>  - repository: adapt `repo_clear()` to fully reset the repository
>  - builtin/clone: don't apply "core.sharedRepository" to leading dirs
>  - builtin/init: move handling of "core.sharedRepository" into "setup.c"
>  - builtin/init: refactor messy creation of leading directories
>  - path: introduce `safe_create_leading_directories_no_share_const()`
>  - path: drop useless `safe_create_leading_directories_1()`
>  - Merge branch 'ps/odb-alternates-at-creation' into ps/setup-enforce-repo-passed-to-create-repository-has-no-state
> 
>  The repository initialization sequence has been refactored to treat
>  the repository object passed to create_repository() purely as an
>  out-parameter.
> 
>  Waiting for response.
>  cf. <b0ec2ef9-7aef-4f7d-b31b-7141b39c2d24@gmail.com>
>  cf. <xmqqtsn9mkog.fsf@gitster.g>
>  cf. <CAOLa=ZSX0e25wK5qQwznXN9rVM+WHn8631pkTEN9Zm-BrXfEsg@mail.gmail.com>
>  source: <20260928-pks-create-repository-stateless-v2-0-a03612f703fa@pks.im>

Hm. All I can see is the discussion around free(3) vs free(3p). Is the
expectation to do a reroll with s/free(3p)/free(3)?

Patrick