| 27 Dec 2025 |
|  zimward changed their display name from zimward @ 39c3 to zimward @ 39c3 ☎️ 75947. | 10:43:52 |
|  jappie changed their display name from jappie @ 39c3 to jasper @ 39c3 ☎️ 62749. | 13:30:50 |
|  matthewcroughan changed their display name from matthewcroughan to matthewcroughan @ 39c3 (DECT 94667 or 97340 or 67192). | 14:45:37 |
 Grimmauld (any/all) | System is now running on staging-next. No obvious issues, other than needing to fetch https://src.fedoraproject.org/rpms/med/raw/ad90858d17f15168db905b095346cd9671de88ea/f/med-gcc15.patch for medfile | 15:27:44 |
| Grimmauld (any/all) | i can PR that, though medfile also has a broken fetcher and i am not sure what to do about that | 15:28:11 |
 K900 | What the fuck is medfile | 15:28:45 |
| K900 | That code looks very cursed and Google gives me Polish medical document management systems | 15:28:45 |
| Grimmauld (any/all) | a dependency to salome smesh, which is a dependency to freecad | 15:29:18 |
| Grimmauld (any/all) | well, freecad has a vendored smesh | 15:29:29 |
| Grimmauld (any/all) | fetcher is broken because cloudflare-like firewall so curl doesn't work. Debian and arch fix this by importing the tarball into salsa/github respectively | 15:30:30 |
| K900 | Amazing | 15:33:07 |
 @ghpzin:envs.net | You can use user agent to work around it:
https://github.com/bmwiedemann/openSUSE/blob/08303e6e850f0de37bfabbd184dae73009f3038b/packages/m/med-tools/med-tools.spec#L32
Seems to work with curl, thought not sure if curlOpts are working properly after latest fetchurl changes. | 15:40:59 |
| @ghpzin:envs.net | You can use user agent to work around it:
https://github.com/bmwiedemann/openSUSE/blob/08303e6e850f0de37bfabbd184dae73009f3038b/packages/m/med-tools/med-tools.spec#L32
Seems to work with curl, thought not sure if curlOptsList is working properly after latest fetchurl changes. | 15:41:49 |
| Grimmauld (any/all) | time to find out | 15:42:27 |
| Grimmauld (any/all) | curl opts seems to work | 15:48:48 |
| Grimmauld (any/all) | https://github.com/NixOS/nixpkgs/pull/474599
There, last PR to make my system run against current staging-next. I hate CAD programs. | 17:51:35 |
| K900 | Extreme nit: does it work with UA set to something like "Firefox (really Nixpkgs, see https://github.com/NixOS/nixpkgs/pull/474599)" or something? | 17:53:41 |
| Grimmauld (any/all) | 
Download image.png | 17:56:43 |
| Grimmauld (any/all) | seems to work, can do | 17:56:49 |
| Grimmauld (any/all) | done | 17:57:31 |
| 28 Dec 2025 |
|  kalbasit joined the room. | 04:44:34 |
| kalbasit | Seems like the last time staging-next was merged into master was 11/14, is there a plan to do one soon? (apologies if I missed an earlier message/context about this...)
Currently, 25.11 has Go 1.25.5 while master has Go 1.25.4 and was hoping to get 1.25.5 to master soon (hitting some blockers on my end)
| 04:47:40 |
 whispers (it/fae) | open PR is https://github.com/NixOS/nixpkgs/pull/471587 | 04:57:03 |
| whispers (it/fae) | last i checked, it might be a minute before merge since python 3.14 and gcc 15 have both caused non-insignificant breakage | 04:58:12 |
| whispers (it/fae) | * | 05:07:45 |
| whispers (it/fae) | * | 05:10:36 |
| kalbasit | I see, makes sense; Thanks! I ended up working around my blocker for now but will keep an eye on this one... | 05:21:26 |
| Grimmauld (any/all) | Tbh there will be a point when we should just merge, breakage be damned. Let maintainers fix their shitbwhen it is broken on master. Five people doing staging (okay maybe ten, it was more people than usual this time) doing fixes for all of nixpkgs isn't viable. | 08:31:30 |
| Grimmauld (any/all) | * Tbh there will be a point when we should just merge, breakage be damned. Let maintainers fix their shit when it is broken on master. Five people doing staging (okay maybe ten, it was more people than usual this time) doing fixes for all of nixpkgs isn't viable. | 08:31:45 |
 leona | it would be good if we could get numpy (on 3.14) and uvloop (3.13 and 3.14) to work on darwin | 08:33:14 |
