v2.51.0.vfs.0.0

[Johannes Schindelin (dscho)]

What's Changed

• Update to Git v2.51.0 by Johannes Schindelin (@dscho) in Rebase to v2.51.0 #785

Full Changelog: v2.50.1.vfs.0.1...v2.51.0.vfs.0.0

---

This discussion was created from the release v2.51.0.vfs.0.0.

[Johannes Schindelin (dscho)]

For the record, I have not forgotten about v2.51. There's just... developments.

First of all, we identified some mistakes I made during the -rc1 rebase, and I wanted to make extra sure that we had found and fixed all of those.

But there is a bigger issue unfolding: There has been a change in upstream Git v2.51 that frequently leads to bogus .lock file issues when cloning/fetching onto a case-insensitive filesystem when some refs differ only in lower-/upper-case letters (which is unfortunately all too common). I am monitoring the development of fixes for this issue and will update this here discussion once I have a better grasp on the likely timeline.

[Johannes Schindelin (dscho)]

While v2 of that patch series landed on Monday, the first reviews that trickled in let me expect a v3 to land (hopefully still this week).

In other news, I backported a fix for another v2.51 regression here: #796 (I consider this also a blocker that needs to be merged before we can release Microsoft v2.51).

[Johannes Schindelin (dscho)]

Well, v3 didn't do it (surprise! </s>). At least for v4, the Git maintainer "declared victory". 6 days after that declaration, though, it still has not been merged to Git's equivalent of the development branch (read: this commit still reports to be only in the seen branch, but not promoted to the next branch), so...

Even for Git's standards, this is quite the dragged-out bug fix process, I have to admit.

Maybe I should just go ahead and merge #798 and deal with the fall-out of merging "early". Wouldn't be the first time 😁

[Johannes Schindelin (dscho)]

Okay, there is finally an official release of Microsoft Git v2.51!