| 15 Nov 2024 |
 hexa | I wish | 21:47:08 |
| hexa | But maybe we'll get there soon | 21:47:20 |
 ma27 | re nix-eval-jobs: what's ypur stance on constituents?
I hacked a patch together for the lix fork, but discussion about what to do with it is still pending. | 22:32:17 |
| 16 Nov 2024 |
|  Viorel-Cătălin Răpițeanu changed their display name from @catalin:one.ems.host to Viorel-Cătălin Răpițeanu. | 18:11:31 |
| 17 Nov 2024 |
 tomberek | In reply to @Ericson2314:matrix.org
I say better to ask for forgiveness than permission :) Sounds good. Proceed! | 00:52:28 |
| 18 Nov 2024 |
 John Ericson | Anyone want to review https://github.com/NixOS/hydra/pull/1422 ? doing the meson one first since the uncertainty around hydra's constituent jobs | 15:26:42 |
 @rick:matrix.ciphernetics.nl | Will do that tonight when I get home (like somewhere between 19 and 20 UTC+1). Thanks for doing that change! | 15:41:27 |
| John Ericson | thanks Mindavi ! | 15:47:25 |
| @rick:matrix.ciphernetics.nl | Still looking but posted initial findings | 21:27:56 |
| 20 Nov 2024 |
|  Inayet removed their profile picture. | 00:59:41 |
|  Kamilla 'ova joined the room. | 12:55:34 |
| ma27 | In reply to @Ericson2314:matrix.org
Anyone want to review https://github.com/NixOS/hydra/pull/1422 ? doing the meson one first since the uncertainty around hydra's constituent jobs re constituent jobs, I hacked something together for nix-eval-jobs (lix variant), just didn't have the time so far to pursue it further: https://git.lix.systems/lix-project/hydra/issues/11 | 17:28:50 |
| John Ericson | ma27: oh great! | 17:29:08 |
| ma27 | it's kinda ugly because I needed a proof-of-concept, but I have some more cleanups planned. | 17:29:39 |
| John Ericson | ma27: do you know if we use these named constitutent jobs? | 17:30:02 |
| John Ericson | vs the ones that just reference other drvs? | 17:30:15 |
| ma27 | assuming you mean nixpkgs with "we", then yes | 17:30:39 |
| ma27 | see e.g. https://github.com/NixOS/nixpkgs/blob/master/nixos/release-combined.nix#L48 | 17:31:03 |
| ma27 | they do make a difference in fact: at work, we switched to named constituents from drvs and that + more higher parallelism (that was possible because the release-job was cheaper) resulted in an eval-time drop from 15min to 50s. | 17:31:48 |
| John Ericson | woah wat? | 20:58:07 |
| 22 Nov 2024 |
|  Morgan (@numinit) joined the room. | 17:50:54 |
| 23 Nov 2024 |
|  @perchun:matrix.org joined the room. | 22:24:49 |
| @perchun:matrix.org | hey, how can I get data from https://hydra.nixos.org/eval/1810064/builds if it exits with 504 Gateway Time-out? is there an option for paging?
https://github.com/NixOS/hydra/blob/f974891c76e295240017dd7f04d50ecb4b70284e/hydra-api.yaml#L554 | 22:25:11 |
 K900 | Please don't scrape Hydra | 23:01:14 |
| K900 | Why do you need that data? | 23:01:20 |
| 24 Nov 2024 |
| @perchun:matrix.org | i would like to create a new channel based on build length (updates only when there are no packages that are long to build; tired of long updates on unstable and tired of compiling chromium on unstable-small) | 00:16:31 |
| K900 | That's not really practical for a variety of reasons | 07:15:11 |
| 25 Nov 2024 |
| @perchun:matrix.org | Is there at least a way to get building time by derivation path? | 12:53:34 |
| K900 | Not easily | 12:54:06 |
| @perchun:matrix.org | So there is, without making stress for the infra? | 12:55:12 |
