!djTaTBQyWEPRQxrPTb:nixos.org

Nixpkgs Architecture Team

219 Members
https://github.com/nixpkgs-architecture, weekly public meetings on Wednesday 15:00-16:00 UTC at https://meet.jit.si/nixpkgs-architecture54 Servers

Load older messages


SenderMessageTime
22 Jul 2024
@jeremiahs:matrix.orgjeremiahs

So @tomberek:matrix.org had pointed me here to talk about an inconsistency I noticed in how nixos handles warnings and assertions. The declaration can be found here: https://github.com/NixOS/nixpkgs/blob/master/nixos/modules/misc/assertions.nix

I am wondering if these can be changed to be more consistent. I had a few ideas of a new way it could be done
here is the expected usage:

logs = [
 {
  # if false the build would fail as prior
  assertion = true;
 # if false  it should show a warning as prior
  warnings = true;
  message = "";
  }
];

in the example above we merge the functionality of assertions and warnings into the data structure of which assertions or warnings must be present to be valid. in addition this could be added in a way as to not break the existing module system but nudge towards a more consistent data structure in the module system. let me know if this makes sense or not.

18:13:36
@infinisil:matrix.orginfinisil jeremiahs: Check out https://github.com/NixOS/nixpkgs/pull/97023! 20:11:36
@infinisil:matrix.orginfinisilHad to be reverted unfortunately, but this would've been so nice 🥲20:11:51
@infinisil:matrix.orginfinisilThere's a bunch of more recent follow-up PR/issue discussions linked to20:12:34
@infinisil:matrix.orginfinisil * There's a bunch of more recent follow-up PR/issue discussions linked to towards the end20:12:40
@jeremiahs:matrix.orgjeremiahs
In reply to @infinisil:matrix.org
There's a bunch of more recent follow-up PR/issue discussions linked to towards the end
reading now
20:24:44
@jeremiahs:matrix.orgjeremiahs infinisil: feel free to correct my understanding, but its not that it cant be done but more work under the hood is required for this to work as expected. 20:57:54
@infinisil:matrix.orginfinisil jeremiahs: Hmm, it does work as expected, but it also causes other things to not work as expected :P 20:59:48
@infinisil:matrix.orginfinisilI believe a satisfying solution can be figured out, but needs some serious brain power21:00:31
@jeremiahs:matrix.orgjeremiahsunderstood21:20:59
23 Jul 2024
@ezzobirbezziou:matrix.orgEzzobir Bezziou joined the room.08:22:12
@felixalb:feal.no@felixalb:feal.no left the room.11:27:47
24 Jul 2024
@redstone-menace:matrix.orgR̴̨͕͇͍̞̮̐̅͆̌̀̉̐͋̈́̃̀͒́̎̅̚̚̚͠͝Ĕ̵̡̛͖͖̟̙̫̱͈̘̞̭͍͍͑̌̄͑̓̋̓̀̈̏̈́͊̇͊͆̉͂̏̀̃̚͘͝͝ͅͅD̶̡̢͔̱̖̮͙͉̘̺͓͍̩̮͈͍͗̃̀̏͌͘͜ͅŚ̸̬̭̯̬͙͇͓̬̩̳̤͚͓̤̩̺͉͖̉͛̓̿̎͊̿̆́̐͂̇͌̄̇̓͘ͅͅT̴̞̫̘̝͇͔̟̪̪̦͂̔̎̀̎ͅŎ̷̡̬̹̪͈̭̣͈̭̭͉̦̖̝̘̪͖͔̥̦̘̻̳Ṋ̶̛̫͈̳̘͚̜̔̋͆̅̈́͊̑͊̉̌̈́̾͑̈́̚ͅË̸̡̨̨̛͇̜̖͔͖̻̟̗̠̙͓̘̗̥͉͇̜͑͆͊͑͑̀̓͒͜͝͝ changed their display name from redstone-menace to Redstone.10:16:19
25 Jul 2024
@willpower3309:matrix.org@willpower3309:matrix.org left the room.15:09:31
@philiptaron:matrix.orgPhilip Taron (UTC-8)

I have a set of PRs related to pkgs/top-level/release-attrpaths-superset.nix and the __attrsFailEvaluation attributes sprinkled through nixpkgs. I'd appreciate any reviews and merging.

They're all linked to https://github.com/NixOS/nixpkgs/issues/324619.

23:40:23
26 Jul 2024
@infinisil:matrix.orginfinisil Philip Taron (UTC-8): Tested and merged, except for the two with merge conflicts :) 00:07:02
@philiptaron:matrix.orgPhilip Taron (UTC-8)Thank you so much!!!00:07:29
@infinisil:matrix.orginfinisilBtw I think a bunch of these could've been done in the same PR with multiple commits, makes it easier to test00:07:31
@infinisil:matrix.orginfinisil

I used

for b in $(git branch --remote --list 'philiptaron/__attrsFailEvaluation/*'); do
  git merge "$b"
done
00:08:03
@infinisil:matrix.orginfinisilTo test all of them together :P00:08:08
@philiptaron:matrix.orgPhilip Taron (UTC-8)I’m gun shy on that: I’ve gotten so much “I can’t review this” negative feedback on “treewide” changes00:08:15
@philiptaron:matrix.orgPhilip Taron (UTC-8)They originated as a single branch!00:08:33
@infinisil:matrix.orginfinisilHaha00:08:39
@infinisil:matrix.orginfinisilI see, no worries ;)00:08:51
@infinisil:matrix.orginfinisilI usually want to err on the side of "merge fast and often", which I guess does work better with smaller PRs, but it's hard to strike the balance00:10:06
@philiptaron:matrix.orgPhilip Taron (UTC-8)Don't I know it. I also like merge fast and often, but, uh, that relies on fast reviews and merges, which are only some of the time present (for obvious open source reasons!)00:11:01
@infinisil:matrix.orginfinisilYeah 😅00:11:55
@infinisil:matrix.orginfinisil Philip Taron (UTC-8): Hmm weird, https://github.com/NixOS/nixpkgs-check-by-name/blob/main/.github/workflows/after-release.yml didn't run 23:29:38
@infinisil:matrix.orginfinisilhttps://github.com/NixOS/nixpkgs-check-by-name/actions/workflows/after-release.yml23:29:47
@philiptaron:matrix.orgPhilip Taron (UTC-8)I saw that. The release went out, though./23:30:09

Show newer messages


Back to Room ListRoom Version: 9