Nix / Nixpkgs / NixOS Dev

388 Members
(#nixos for questions) | NixOS 21.05 Okapi ✨ https://nixos.org/blog/announcements.html#21.05 | https://hydra.nixos.org/jobset/nixos/trunk-combined https://channels.nix.gsc.io/graph.html | https://r13y.com | 20.09 RMs: worldofpeace, jonringer90 Servers

Load older messages

22 Oct 2021
@vcunat:matrix.orgVladimír ČunátI don't know LLVM stuff, so I'm more likely to be wrong.16:55:50
@qyliss:fairydust.spaceAlyssa RossIdeally we'd just have a single version of LLVM for everything16:56:48
@toonn:matrix.orgtoonn Maybe that info also predated the branching monstrosity? 16:56:55
@qyliss:fairydust.spaceAlyssa RossYeah it did16:57:06
@toonn:matrix.orgtoonn I do agree but that's such a massive job to update then. 16:57:22
@qyliss:fairydust.spaceAlyssa Rossbut we should be trying to eliminate the branching, not introduce even more variance16:57:38
@toonn:matrix.orgtoonn My PR does eliminate one branch though. 16:58:00
@toonn:matrix.orgtoonn I'm aligning the Darwin platforms. 16:58:14
@qyliss:fairydust.spaceAlyssa Rossdepends how you look at it16:58:16
@qyliss:fairydust.spaceAlyssa Rossyou're aligning Darwin, but making x86_64 out of sync16:58:25
@toonn:matrix.orgtoonn True. 16:58:37
@qyliss:fairydust.spaceAlyssa RossBut look, if the result of this is that in future we can update LLVM for Linux without having to wait for Darwin to catch up, that's fine with me, as long as we all agree16:59:12
@vcunat:matrix.orgVladimír ČunátEither way, once you're confident about (how to do) this, I can configure jobset(s). Hydra's not even busy currently.17:01:11
@qyliss:fairydust.spaceAlyssa RossLet's just go ahead with toonn's PR, and make a new one for updating it on Linux.17:02:01
@qyliss:fairydust.spaceAlyssa RossI don't want to derail toonn's efforts at the last minute17:02:09
@toonn:matrix.orgtoonn If you give me a Linux jobset I can try just bumping the default on Linux and seeing whether there's fallout. 17:02:16
@qyliss:fairydust.spaceAlyssa RossI just wish we'd had consensus on the process beforehand.17:02:33
@qyliss:fairydust.spaceAlyssa Rosstoonn: let's do a separate PR for it — yours is already quite big and reviewing it will be a big job 17:03:12
@qyliss:fairydust.spaceAlyssa Rossbut maybe base it on your current one?17:03:29
@toonn:matrix.orgtoonn Sure, that's fine. The current one is based on staging-next form a week or two ago actually. Maybe that still needs to be rebased? I wanted to keep it close to the commit I used for the evaluation. 17:04:26
@qyliss:fairydust.spaceAlyssa Rosskeep it as-is as long as it merges imo17:05:04
@qyliss:fairydust.spaceAlyssa Rossit's good if it doesn't change while people are reviewing it17:05:52
In reply to @eyjhb:eyjhb.dk
Sandro: RE https://github.com/NixOS/nixpkgs/pull/141477#discussion_r734447728 , what shoulud I do about that?
Depending on what gets earlier merged we need to change either PR or we get an eval error
@toonn:matrix.orgtoonn Hmm, I just had an EditorConfig Action fail for #142593 but it looks like the runner simply failed to fetch the branch, can Actions be restarted? 17:29:55
@sandro:supersandro.deSandrobutton top right https://github.com/NixOS/nixpkgs/runs/3978618643?check_suite_focus=true 17:37:43
@sandro:supersandro.deSandrobut I think you want to solve that merge conflict to get that working17:37:54
@toonn:matrix.orgtoonn Hmm, problem is we're trying to keep it as close as possible to the Darwin bump and the conflict has been introduced since then. Is ignoring the failed Actions until that has been merged an option? 17:54:41
@sandro:supersandro.deSandrosure but then we can't merge it earlier18:06:39
@fabianhjr:matrix.orgfabianhjrAnybody has an example of flakes + dockerTools? Can't find one from googling around and searching the github org.18:07:01
In reply to @sandro:supersandro.de
Depending on what gets earlier merged we need to change either PR or we get an eval error
Lets see what comes first then. Unsure who should merge the prometheus update sadly....

There are no newer messages yet.

Back to Room List