FinderGit 0.32.0 is out, and this one came entirely from using the thing on real data.
0.31.0 let you select many tags and act on them at once. Worked great until someone opened it on a repo with 715 tags, scrolled down to the release candidates from four years ago, selected three, and found that the bar with every action for that selection was several hundred rows above, off the top of the panel
Someone asked me why v1.0.2 was sitting at the bottom of a 207-tag list instead of between 1.0.1 and 1.0.3. Sorting bug, right?
Nope And chasing it explained three other things they'd asked in the same message.
Git has two kinds of tag. An annotated one is a real object with a message and a creation date. A lightweight one is just a name pointing at a commit git records nothing about when it was made. So the only date it has is the commit's, and that commit was six days older. The list was right and read as though it were broken.
If you keep clones from more than one place, FinderGit used to be uneven about it. GitHub filled in its numbers once you connected an account. Codeberg filled them in with no setup at all. GitLab and Bitbucket showed you nothing.
Now all four are read and the last three need nothing from you. No account, no token.
