| 17 Dec 2025 |
 vcunat | It wasn't a typo, you didn't want this one, right? | 08:35:41 |
 Emil Thorsøe | doesn't staging-next pull in staging-nixos anyway? | 09:02:14 |
| Emil Thorsøe | or like that would sound sensible to me | 09:02:28 |
 K900 | Yes | 09:04:36 |
| vcunat | I do merge staging-nixos whenever merging staging-next to master. (or at least I'm trying to remember) | 09:08:16 |
 emily | maybe we should have automated staging-nixos → staging-next merges | 09:21:08 |
| emily | rather than separate master → staging-next and master → staging-nixos merges | 09:21:18 |
| emily | that way it would happen by itself | 09:21:31 |
| vcunat | So master → staging-nixos → staging-next → staging? 🤔😁 | 09:25:44 |
| vcunat | Making the chain longer does have some risks, but maybe that's just... inherent to our situation. | 09:27:21 |
 Grimmauld (any/all) | its not a terrible idea tbh | 09:33:20 |
| Grimmauld (any/all) | if github buckles we'll just have to move away from github :P | 09:33:37 |
| emily | potentially means resolving a conflict only once rather than thrice, if master, staging-nixos, and staging-next all differ on something | 09:36:32 |
| emily | say, Python stuff: change to test driver on staging-nixos, mass update on staging/staging-next | 09:36:48 |
| emily | and some conflicting non-rebuilding Python packaging change on master | 09:37:07 |
| emily | well, I guess this is still two conflict resolutions to do. but less confusing ones hopefully | 09:37:23 |
| emily | otherwise you resolve the master/staging-nixos and master/staging-next ones separately and then have to deal with staging-nixos/staging-next later | 09:37:45 |
| emily | the whole periodic merge arrangement is not great IMO, especially when we get semantic conflicts, but… that's another matter | 09:38:28 |
| emily | especially having to manually trigger them for follow-up PRs sucks | 09:38:40 |
| K900 | Do we not have the beautifulsoup fix in yet | 13:02:32 |
| K900 | I decided to send one of my configs | 13:02:41 |
| K900 | So far the fallout is LKL (fixed in master, needs bump), timidity (uh oh) and beautifulsoup | 13:02:59 |
 Marcus | my browser is stuggling so hard with newly failing 🙈 | 13:03:04 |
| K900 | GCC upgrades are like that yeah | 13:04:37 |
| Grimmauld (any/all) | we need new bootstrap tools for musl too | 13:22:08 |
| Grimmauld (any/all) | from before the gcc bump | 13:22:23 |
| K900 | Uhh what why | 13:22:43 |
| Grimmauld (any/all) | because bootstrap is dead on gcc 15 rn or something | 13:23:15 |
| Grimmauld (any/all) | dunno but old musl bootstrap tools no longer worked on gcc 15, and on gcc 15 you don't get far enough to build new bootstrap tools without making new gcc 14 musl bootstrap tools | 13:23:54 |
 Fabián Heredia | timidity has a fix since two days ago: https://github.com/NixOS/nixpkgs/pull/470974
Merging | 15:17:06 |
