#nixos-dev on 2021-03-17

2020-10-29 21:47 worldofpeace changed the topic of #nixos-dev to: NixOS Development (#nixos for questions) | NixOS 20.09 Nightingale ✨ https://discourse.nixos.org/t/nixos-20-09-release/9668 | https://hydra.nixos.org/jobset/nixos/trunk-combined https://channels.nix.gsc.io/graph.html | https://r13y.com | 20.09 RMs: worldofpeace, jonringer | https://logs.nix.samueldr.com/nixos-dev

00:12 orivej has quit [Ping timeout: 245 seconds]

00:33 NinjaTrappeur has quit [Ping timeout: 264 seconds]

00:34 cole-h_ has quit [Ping timeout: 265 seconds]

00:41 NinjaTrappeur has joined #nixos-dev

00:49 supersandro2000 has quit [Disconnected by services]

00:49 supersandro2000 has joined #nixos-dev

00:54 rajivr has joined #nixos-dev

01:15 NinjaTrappeur has quit [Ping timeout: 244 seconds]

01:22 NinjaTrappeur has joined #nixos-dev

01:56 ris has quit [Ping timeout: 256 seconds]

02:43 evils has quit [Ping timeout: 246 seconds]

02:50 evils has joined #nixos-dev

03:22 teto has quit [Ping timeout: 272 seconds]

03:48 s1341_ has quit [Quit: Connection closed for inactivity]

03:55 <siraben> tazjin: that's crazy, imagine the amount of conflicts and constant rebases

03:55 <siraben> or do rebases not scale when you have dozens of people working on the same branch

05:16 orivej has joined #nixos-dev

06:08 cole-h_ has joined #nixos-dev

06:17 cole-h_ is now known as cole-h

06:44 srk has quit [Remote host closed the connection]

06:44 srk has joined #nixos-dev

07:14 orivej has quit [Ping timeout: 256 seconds]

07:29 orivej has joined #nixos-dev

07:39 jonringer has quit [Ping timeout: 264 seconds]

07:42 orivej has quit [Ping timeout: 276 seconds]

07:46 orivej has joined #nixos-dev

07:49 pmy has quit [Quit: WeeChat 3.1]

07:49 pmy has joined #nixos-dev

07:51 pmy has quit [Client Quit]

07:52 pmy has joined #nixos-dev

07:55 cole-h has quit [Ping timeout: 264 seconds]

07:58 orivej has quit [Ping timeout: 264 seconds]

08:02 evanjs has quit [Read error: Connection reset by peer]

08:03 evanjs has joined #nixos-dev

08:18 teto has joined #nixos-dev

09:32 <tazjin> siraben: everything is (conceptually) rebase only at Google, and your unit of work is always one commit equivalent - I'd suspect Uber is the same. This means the problem doesn't come up for the most part

09:51 pmy_ has joined #nixos-dev

09:51 evils_ has joined #nixos-dev

09:55 pmy has quit [*.net *.split]

09:55 evils has quit [*.net *.split]

10:18 devhell has joined #nixos-dev

10:25 __monty__ has joined #nixos-dev

10:45 Raito_Bezarius has quit [Ping timeout: 264 seconds]

10:54 evils_ has quit [Ping timeout: 264 seconds]

10:56 mkaito has joined #nixos-dev

10:57 Raito_Bezarius has joined #nixos-dev

11:11 mkaito has quit [Quit: WeeChat 3.1]

11:15 mkaito has joined #nixos-dev

11:16 Synthetica has joined #nixos-dev

11:44 orivej has joined #nixos-dev

11:50 AlwaysLivid has joined #nixos-dev

11:52 pmy_ has quit [Quit: WeeChat 3.1]

11:57 orivej has quit [Ping timeout: 264 seconds]

11:59 pmy has joined #nixos-dev

12:00 orivej has joined #nixos-dev

12:04 orivej has quit [Ping timeout: 265 seconds]

12:14 orivej has joined #nixos-dev

12:18 mkaito has quit [Read error: Connection reset by peer]

12:30 orivej has quit [Ping timeout: 246 seconds]

12:34 pmy has quit [Quit: WeeChat 3.1]

12:36 pmy has joined #nixos-dev

12:36 pmy has quit [Client Quit]

13:19 orivej has joined #nixos-dev

13:23 mkaito has joined #nixos-dev

13:23 mkaito has quit [Changing host]

13:24 orivej has quit [Ping timeout: 256 seconds]

13:43 tv has quit [Ping timeout: 245 seconds]

13:47 tv has joined #nixos-dev

13:52 orivej has joined #nixos-dev

14:08 adisbladi is now known as adisbladis

14:17 <gchristensen> so Nix technically allows for packages to have `.`s in attribute names

14:17 <gchristensen> should Nixpkgs?

14:18 orivej has quit [Ping timeout: 256 seconds]

14:18 <siraben> would be confusing

14:18 <siraben> what does `foo.bar.baz` mean?

14:18 <gchristensen> that has a clear meaning(ish), foo."bar.baz" would be how to do it

14:19 <siraben> ah

14:19 <adisbladis> I think the answer is nuanced

14:19 <adisbladis> Which packages are you thinking of where this would be relevant?

14:19 <Synthetica> Doesn't sound like something we want to do on first glance, what's your intended usecase? Version numbers?

14:20 <adisbladis> I tend to think that we should follow upstream names ~110% of the time

14:20 <sterni> gchristensen: it would be a pain to specify nix-build -A \"foo\" on the command line

14:20 <niksnut> gchristensen: no, and also it shouldn't have package names starting with digits

14:20 <gchristensen> usecase? none. I'm looking at some deep dark Hydra code :)

14:20 <niksnut> nix-env filters them out

14:20 <gchristensen> hmm yeah

14:21 <jtojnar> kill nix-env then

14:21 <niksnut> IIRC hydra also filters out bad attribute names

14:21 <sterni> adisbladis: we probably should make an effort to get closer to that but not sure if dots contained is the place to start :p

14:21 <niksnut> because dots definitely break the project.jobset.job notation

14:21 <gchristensen> for sure, right

14:21 <sterni> niksnut: attribute names starting with a number are a parse error though, right?

14:22 <niksnut> not if you quote them

14:22 <gchristensen> no: nix-repl> { "9foo.bar" = "baz"; }

14:22 <gchristensen> { "9foo.bar" = "baz"; }

14:22 <adisbladis> jtojnar++

14:22 <{^_^}> jtojnar's karma got increased to 0b1000110

14:22 <niksnut> probably we never should have added quoted attrs

14:22 <sterni> ah right, yeah

14:23 <sterni> they are pretty useful

14:26 teto has quit [Ping timeout: 244 seconds]

14:26 jonringer has joined #nixos-dev

14:26 <LinuxHackerman> > {"" = null;}

14:26 <{^_^}> { "" = null; }

14:27 <LinuxHackerman> hehehe

14:28 <sterni> > {"" = 42;}.""

14:28 <{^_^}> 42

14:28 <ajs124> > { "Ö" = "ä"; }

14:28 <{^_^}> { "Ö" = "ä"; }

14:28 <ajs124> > { "💯" = "🔚"; }

14:28 <{^_^}> { "💯" = "🔚"; }

14:29 <gchristensen> oh no

14:29 <sphalerite> Agree that they're useful though. Without them it would be a lot harder to produce arbitrary JSON from nix.

14:30 <adisbladis> sphalerite: And maybe more importantly, to consume arbitrary json

14:30 <sphalerite> oh yeah.

14:30 <adisbladis> gchristensen: It breaks nix repl :D

14:30 <gchristensen> what does?

14:31 <siraben> > (x: x x) (x: x x)

14:31 <etu> > "Räksmörgås"

14:31 <{^_^}> error: stack overflow (possible infinite recursion)

14:31 <adisbladis> gchristensen: Using emoji

14:31 <{^_^}> "Räksmörgås"

14:31 <siraben> sorry not sorry

14:31 <gchristensen> oh hehe

14:31 <sterni> adisbladis: yeah nix repl has some issues with characters with wcwidth > 1 (presumably)

14:32 <adisbladis> > { "🌮" = hello; }

14:32 <{^_^}> { "🌮" = <CODE>; }

14:32 <sterni> as long no one finds a encoding-related exploit for string attrsets we are fine I guess :p

14:33 <sterni> > { "foobar" = 12;}."foo‍bar"

14:33 <{^_^}> attribute 'foo‍bar' missing, at (string):489:1

14:33 <siraben> > "foo\0bar"

14:33 <{^_^}> "foo0bar"

14:33 <sterni> siraben: NUL byte is the only byte not allowed in nix strings

14:35 <siraben> How do you convert an int to a char in Nix?

14:35 <ajs124> siraben: fromJSON

14:36 <sterni> siraben: https://code.tvl.fyi/tree/users/sterni/nix/char/default.nix#n27

14:36 <sterni> (via a lookup table essentially)

14:37 <siraben> Ah, but no way to workaround the nul byte

14:37 <siraben> heh

14:37 <siraben> wanted to do lib.ord 0

14:37 <siraben> I mean lib.chr 0

14:39 <sterni> nul byte is strictly off limits

14:44 <siraben> aw shucks

15:09 rj_ has joined #nixos-dev

15:14 orivej has joined #nixos-dev

15:17 s1341_ has joined #nixos-dev

15:42 rng4 has joined #nixos-dev

15:45 teto has joined #nixos-dev

15:49 b42 has quit [Ping timeout: 264 seconds]

15:50 b42 has joined #nixos-dev

16:03 orivej has quit [Ping timeout: 256 seconds]

16:03 orivej has joined #nixos-dev

16:10 cole-h has joined #nixos-dev

17:19 devhell has quit [Quit: leaving]

17:37 teto has quit [Ping timeout: 256 seconds]

17:40 <rmcgibbo[m]> Does anyone understand why hydra would show "This job is not a member of the latest evaluation of its jobset. This means it was removed or had an evaluation error." for a package on aarch64-linux (but not x86_64-linux)? But (a) I don't actually see an evaluation error in the tab that shows all of the evaluation errors, and (b) ofborg didn't throw an evaluation error either, and showed it as a changed attr.

17:40 <rmcgibbo[m]> https://github.com/NixOS/nixpkgs/pull/116641#issuecomment-801274731

17:51 teto has joined #nixos-dev

17:59 teto has quit [Ping timeout: 264 seconds]

17:59 <drakonis> what exactly is the structured-attrs branch on nixpkgs?

18:01 rajivr has quit [Quit: Connection closed for inactivity]

18:03 <LinuxHackerman> drakonis: it enables __structuredAttrs in stdenv

18:03 <LinuxHackerman> globin wrote a nice blog post about it https://nixos.mayflower.consulting/blog/2020/01/20/structured-attrs/

18:03 <drakonis> fantastic.

18:11 teto has joined #nixos-dev

18:24 ris has joined #nixos-dev

18:31 teto has quit [Ping timeout: 256 seconds]

18:56 AlwaysLivid has quit [Remote host closed the connection]

19:22 AlwaysLivid has joined #nixos-dev

19:30 drakonis has quit [Quit: ZNC 1.8.2 - https://znc.in]

19:32 drakonis has joined #nixos-dev

19:44 thonkpod_ is now known as thonkpod

20:02 Baughn has quit [Ping timeout: 265 seconds]

20:24 orivej has quit [Ping timeout: 264 seconds]

20:28 evils has joined #nixos-dev

20:38 orivej has joined #nixos-dev

21:07 rj_ has quit [Ping timeout: 268 seconds]

21:12 teto has joined #nixos-dev

21:18 <supersandro2000> Are such throws permitted https://github.com/NixOS/nixpkgs/pull/116126/files#diff-b1708e8e6efcd6d64053178a991c4143fc58a6497a87f122289deec75fce7e23R54 ?

21:34 <samueldr> if the tmux package set didn't "recurse into" previously, or right now, it wouldn't cause issues... but it's probably not good?

21:43 Jackneill has quit [Ping timeout: 246 seconds]

22:05 <supersandro2000> that is the commit that removed mkDerivation https://github.com/nprindle/nixpkgs/commit/daea4b1134f46c39a886143a0cbcb55d95d54529#diff-b1708e8e6efcd6d64053178a991c4143fc58a6497a87f122289deec75fce7e23

22:14 <sterni> rmcgibbo[m]: unfree / unsupported platform are also eval errors which hydra can deal with

22:15 <sterni> rmcgibbo[m]: if a package isn't available for aarch64 it'd just be filtered out

22:17 justanotheruser has quit [Quit: WeeChat 2.9]

22:17 <sterni> but not sure what exactly is happening in that case

22:22 Baughn has joined #nixos-dev

22:27 justanotheruser has joined #nixos-dev

22:30 orivej has quit [Ping timeout: 256 seconds]

22:34 Baughn has quit [Remote host closed the connection]

22:35 Baughn has joined #nixos-dev

22:39 orivej has joined #nixos-dev

23:02 orivej has quit [Ping timeout: 245 seconds]

23:03 Baughn has quit [Ping timeout: 260 seconds]

23:09 <rmcgibbo[m]> I guess maybe in general I don't understand how to find stuff on hydra.

23:09 <rmcgibbo[m]> nix-shell -p hydra-check --command 'hydra-check _1password-gui --url --channel master' returns https://hydra.nixos.org/job/nixpkgs/trunk/_1password-gui.x86_64-linux

23:09 <rmcgibbo[m]> which doesn't give anything.

23:09 <rmcgibbo[m]> but https://github.com/NixOS/nixpkgs/pull/116675 is updating it

23:09 <{^_^}> #116675 (by r-ryantm, 1 hour ago, open): _1password-gui: 8.0.27 -> 8.0.28

23:10 <rmcgibbo[m]> oh no, maybe that's because it's unfree as you said @sterni. actually it makes perfect sense.

23:11 <ryantm> That would explain it.

23:12 xwvvvvwx has quit [Ping timeout: 264 seconds]

23:15 Baughn has joined #nixos-dev

23:16 Baughn has quit [Client Quit]

23:16 Synthetica has quit [Quit: Connection closed for inactivity]

23:16 Baughn has joined #nixos-dev

23:20 xwvvvvwx has joined #nixos-dev

23:36 <supersandro2000> Build Status for _1password-gui.x86_64-linux on master

23:36 <supersandro2000> ⚠ This job is not a member of the latest evaluation of its jobset. This means it was removed or had an evaluation error

23:47 __monty__ has quit [Quit: leaving]