| 24 Mar 2025 |
 ElvishJerricco | * https://github.com/NixOS/nixpkgs/pull/392904 | 23:19:20 |
| ElvishJerricco | Tristan Ross, emily: I wonder if it's actually bad for me to revert the revCount though. I think, as nixpkgs moves forward in time, that number is meant to be monotonic. If, for some reason, anyone has been using it as an ever-increasing version number (even though it was the same number for some 5 or 6 years by 24.11), then this would bring the number down and cause a problem | 23:33:18 |
 emily | probably just leave it increased | 23:33:43 |
| emily | since it's been like that for half a year now without issue | 23:33:47 |
| ElvishJerricco | yea that's what I'm thinking | 23:33:51 |
 Tristan Ross | In reply to @elvishjerricco:matrix.org
Tristan Ross, emily: I wonder if it's actually bad for me to revert the revCount though. I think, as nixpkgs moves forward in time, that number is meant to be monotonic. If, for some reason, anyone has been using it as an ever-increasing version number (even though it was the same number for some 5 or 6 years by 24.11), then this would bring the number down and cause a problem I think the flake changes the rev as well so I would go with Emily's suggestion for non flake users who uses it. | 23:34:39 |
| ElvishJerricco | I will revert the shortRev though, since the old one appears to not be a real rev? While the new one is | 23:34:39 |
| Tristan Ross | Huh | 23:34:53 |
| ElvishJerricco | or, emily is there a fancy jj I could do to make sure it's not just some ambiguous prefix of a rev? | 23:36:03 |
| ElvishJerricco | actually no I think git would say it's ambiguous, not "bad revision 'gfedcba'" | 23:36:23 |
| emily | there's a very easy way to tell that gfedcba isn't a real rev | 23:36:30 |
| Tristan Ross | git show? | 23:36:44 |
| emily | (but also the g is a prefix, not part of the actual hash) | 23:36:46 |
| emily | (hint: python3 -c 'print("".join(reversed("fedcba")))') | 23:37:20 |
