!UNVBThoJtlIiVwiDjU:nixos.org

Staging

393 Members
Staging merges | Running staging cycles: https://github.com/NixOS/nixpkgs/pulls?q=is%3Apr+is%3Aopen+head%3Astaging-next+head%3Astaging-next-25.11 | Review Reports: https://malob.github.io/nix-review-tools-reports/125 Servers

Load older messages


SenderMessageTime
1 Jul 2026
@emilazy:matrix.orgemilyand those are blocking lots of things per the reports (https://malob.github.io/nix-review-tools-reports/)15:52:02
@emilazy:matrix.orgemilythanks for taking up the whack-a-mole here 🫠15:52:11
@emilazy:matrix.orgemilyat least this is good prep for switching to LLD by default!15:52:22
@emilazy:matrix.orgemilyalso, whoa, Hydra shows colours in build logs on the web now15:52:31
@k900:0upti.meK900Yeah that was a funny side effect of the new queue runner15:53:16
@k900:0upti.meK900And some very shitty javascript on top15:53:21
@lach:delta.rocksLachGotta love how numpy deprecates things the way that results in hard to debug NaNs in downstream code D:16:09:51
@hexa:lossy.networkhexaI like numpy's maintenance model a lot16:10:20
@hexa:lossy.networkhexaour maintainer just doesn't read the release notes and tests very little16:10:34
@winter:catgirl.cloudWinterare you volunteering? ;)16:11:23
@lach:delta.rocksLachThis time it was code written by us and it had no deprecation warnings, its just unfortunate behavior change16:11:34
@hexa:lossy.networkhexaI'm fucking volunteering for all of python already, am I not?16:12:53
@sempiternal-aurora:matrix.orgMyriaDo we have a current plan for this? I was mildly interested in why we don't use it already16:13:48
@grimmauld:m.grimmauld.deGrimmauld (any/all) looking at the rebuild counts, please tell me this isn't gst 16:14:22
@grimmauld:m.grimmauld.deGrimmauld (any/all)okay its not16:16:08
@grimmauld:m.grimmauld.deGrimmauld (any/all)my guess is on valkey bump then16:16:25
@lach:delta.rocksLachI wonder what to do with https://github.com/numbagg/numbagg btw... It wasn't updated to numpy 2.5, and the patch isn't small, would it be ok to have such update in nixpkgs?.. It is only used by one package and it is an optional dependency here, but it breaks other packages when I disable it16:18:02
@matthewcroughan:defenestrate.itmatthewcroughan Why is lkl all of a sudden broken on musl but wasn't previously? 16:20:31
@matthewcroughan:defenestrate.itmatthewcroughan unknown type name ‘dev_t’; did you mean ‘div_t’?
unknown type name ‘mode_t’
16:21:31
@hexa:lossy.networkhexastart by filing an issue16:21:36
@matthewcroughan:defenestrate.itmatthewcroughanLike why would this have worked in the past but not now?16:21:42
@hexa:lossy.networkhexais this even related to staging?16:21:58
@matthewcroughan:defenestrate.itmatthewcroughan Or is lkl now included in the build time closure for pkgs.nixos {} now and it wasn't before? 16:21:58
@matthewcroughan:defenestrate.itmatthewcroughanWhenever something is newly broken on musl I post here, should I post somewhere else instead?16:22:26
@matthewcroughan:defenestrate.itmatthewcroughanI track staging in order to stay up to date on musl specific breakage16:22:50
@matthewcroughan:defenestrate.itmatthewcroughan ah yeah lkl is used for nftables which I had switched to, to avoid issues in the past, but now lkl is busted on musl this causes an issue, my bad 16:23:53
@emilazy:matrix.orgemilyI think the plan is just that Randy is going to flip the switch17:02:54
@emilazy:matrix.orgemilyor anyone else could if they wanted to work on it17:02:59
@emilazy:matrix.orgemily there was previously an issue with some stuff using the -r flag in bootstrap which LLD doesn't implement or something, but it's been a while since anyone tried 17:03:17
@emilazy:matrix.orgemily and we can always have people add ld64 as a dep to packages that need it at a pinch if necessary, though of course it'd be nice to phase out the unmaintained linker entirely at some point 17:03:46

Show newer messages


Back to Room ListRoom Version: 6