Re: emacs.git 'master' reset
[email protected] Sun, 26 Jul 2026 16:53:44 +0200
| Newsgroups | gmane.emacs.devel |
|---|---|
| Message-ID | <[email protected]> |
On Sun, Jul 26, 2026 at 05:48:15PM +0300, Eli Zaretskii wrote: > > Date: Sun, 26 Jul 2026 16:31:11 +0200 > > From: [email protected] > > Cc: [email protected], [email protected] > > > > On Sun, Jul 26, 2026 at 04:20:26PM +0300, Eli Zaretskii wrote: > > > > [...] > > > > > If you seriously suggest that everyone who has a local clone of > > > emacs.git should do the above dance, then I find it hard to believe, > > > and hope it is very far from reality. > > > > This local dance is only necessary when one commits to a "main" branch > > while working along (main meaning that there are other people outside > > doing the same). > > > > What I usually do is to work in local work branches (and possibly > > publishing them, depending on whether others want to see what I'm > > up to). > > > > Only when something is agreed upon to go into such a main branch it > > is merged (and pushed, as quickly as possible) to avoid the kind of > > situation you described. > > > > In bigger projects, there's only one person (or a closely-knit group) > > who is supposed to touch those "main" branches, and typically /only/ > > by merging. > > Is this really relevant to the case in point? The case in point is > people who, like me, have the master branch checked out in a local > clone of emacs.git, and that local clone is currently in a state it > was before the upstream was reset to before the commit which mentioned > Claude. There are no other complications. Ah -- so you were in sync with the old master before the change? That's fine, then. I misread you to say you had some commits which hadn't been pushed, sorry for that. Yes, if the git gurus did it right, you get all commits except the problematic one: the original instructions should "just work". Note that they get different hashes, because a commit's hash is chained with previous commit's hashes, so if one is missing (or otherwise changed) the whole history from then on gets different hashes. Cheers -- t
signature.asc
(application/pgp-signature, 195 B)
-----BEGIN PGP SIGNATURE----- iF0EABECAB0WIQRp53liolZD6iXhAoIFyCz1etHaRgUCamYfcQAKCRAFyCz1etHa RlVHAJ9jQWQxIQb2rA8dHJpfMuu6V2QEAwCbBod3pIi3kLM6XCcebzUZR0FdOd8= =8117 -----END PGP SIGNATURE-----