>>>>> On Wed, 21 Apr 2021 14:17:37 +0200, "H.Merijn Brand" <perl5@tux.freedom.nl> said: > On Wed, 21 Apr 2021 07:56:39 -0400, "Ricardo Signes" > <perl.p5p@rjbs.manxome.org> wrote: >> On Wed, Apr 21, 2021, at 3:41 AM, Andreas Koenig wrote: >> > The release has been tagged with v5.33.9 but the tag is a >> > lightweight tag, not an annotated one. This is what the release >> > manager guide has to say about the tag: >> > >> > % grep 'git tag v' Porting/release_managers_guide.pod >> > $ git tag v5.11.0 -m 'First release of the v5.11 series!' >> >> I'm not sure why we prefer an annotated tag, but consistency *is* >> nice. >> >> We can forcibly replace that tag with a new one. Some remotes will >> still have the old one, unless they delete and re-fetch, but I'm >> inclined to think that there's no harm in doing this. >> >> Andreas? Anybody else? I wouldn't suggest to repair this concrete instance. 5.33.7 was also done wrong and many before 5.10 were also not annotated. But >350 times it has been done right and we have now in the last ten years exactly two slips, 5.33.7 and 5.33.9. > I see no added values in comments/annotations on release tags unless > they actually have to add value in the description. As with the above > example, it does not IMHO The output of git describe varies with annotated vs not annotated tags. I don't know who else relies on the distinction. -- andreasThread Previous | Thread Next