| 1 Jul 2026 |
 hexa | it is in there | 02:18:21 |
| hexa | I did the merge | 02:18:24 |
 emily | on the bright side we're only ~4d into the cycle | 02:19:31 |
| emily | probably still going to be faster than the before times | 02:19:48 |
| hexa | I remember staging cycles taking a week to ten days :p | 02:20:55 |
| hexa | good times | 02:21:01 |
| hexa | python-updates also looked fairly different | 02:21:11 |
| hexa | it's hard to describe how much nixpkgs has changed | 02:21:28 |
| hexa | ask vcunat, he's probably with nixpkgs for the last 15y | 02:21:53 |
| emily | whenever I look back at PRs etc. from 2019 when I first got involved it seems like another world | 02:22:33 |
| emily | but even 23.xx era feels very different to now | 02:22:53 |
| hexa | 100% that | 02:22:54 |
| hexa | these days when I work on other projects I miss all the ci functions we have | 02:23:29 |
| emily | meanwhile I'm always feeling despair over our tooling being anaemic :) | 02:38:21 |
| emily | looks like we lost about ~67k successes and ~68k failures 😖 | 04:46:42 |
| emily | Randy Eckenrode: MoltenVK is broken, looks like the same ld64 nonsense again 🫠| 04:47:50 |
| emily | I have a feeling downstream users are going to be seeing a fair number of failing builds thanks to that until next cycle | 04:49:30 |
 Grimmauld (any/all) | In reply to @emilazy:matrix.org
but even 23.xx era feels very different to now A big part of that is fmt too | 05:58:10 |
| Grimmauld (any/all) | It just looks so different and non-standardized | 05:58:40 |
| emily | yes, everything looks worse now :P | 06:10:22 |
| emily | (no, I don't mean it, the nixfmt style is better in many ways. but worse in enough of them to still make me grumble sometimes…) | 06:10:50 |
| emily | it's definitely a shock to look back at some of the very ad-hoc formatting some files had back then and remember what it was like | 06:11:11 |
| Grimmauld (any/all) | occasionally i come accross eelco-era formatting in blames, where the arguments to packages are all clumped together at the top.... Always feels so wrong. | 06:12:30 |
| emily | it was no goddamn fun for conflicts I'll tell you that | 06:12:46 |
 Myria | Has anyone got experience with perl stuff? perlPackages.HTMLForm is failing, which is causing texliveFull build failures, and I've not dealt with perl before. Can give it a shot if no one else wants to investigate, looks like the tests are failing because of a new trailing newline added | 06:51:09 |
| Myria | For reference: https://hydra.nixos.org/build/333325496 | 06:52:45 |
 K900 | Paperless bisect update: let it run overnight, it died to flaky tests in django in under an hour | 07:11:15 |
| K900 | The 3-5 business days bit is really coming back to haunt me huh | 07:11:35 |
 ghpzin | Need update to 6.13 or patches from these:
https://github.com/libwww-perl/HTML-Form/commit/90132a629f2fc77449718491d25e6c5e72805d2c
https://github.com/libwww-perl/HTML-Form/commit/2f738e9762bb8a1929a951ad902f674c56ed7a76
because of perlPackages.HTTPMessage update:
https://github.com/NixOS/nixpkgs/commit/d95a5e6344e6e7b308e6c6e35cadfd37da5229a9 | 07:12:56 |
| Myria | Ah, that's easy to do then, happy to do it unless you want too, but will be a few hours | 07:34:40 |
