| 26 Oct 2023 |
 @adam:robins.wtf | home-manager has similar launchd options, so you could use HM instead | 13:18:04 |
| @adam:robins.wtf | * home-manager has similar launchd options, so you could use HM instead for user timers | 13:18:16 |
 Lily Foster | In reply to @adam:robins.wtf
home-manager has similar launchd options, so you could use HM instead i imagine those are tied to user sessions? | 13:19:49 |
| @adam:robins.wtf | yes, sorry didn't realize this was the ofborg room :) | 13:20:03 |
| @adam:robins.wtf | for GC, nix-darwin has nix.gc options | 13:20:13 |
| Lily Foster | In reply to @adam:robins.wtf
for GC, nix-darwin has nix.gc options i figured it might | 13:20:22 |
| @adam:robins.wtf | https://daiderd.com/nix-darwin/manual/index.html#opt-nix.gc.automatic | 13:20:35 |
 cole-h | In reply to @lily:lily.flowers
i actually do since i have before. if you want me to whip up something. i'm assuming you're just wanting periodic garbage collection? Actually no 😅 The issue is because we’re building (some) x86 and arm Darwin on the same machines | 13:40:27 |
| cole-h | GC is doing fine, it’s the Rosetta stuff that never gets cleaned up unless I manually run a command | 13:40:52 |
| cole-h | In reply to @adam:robins.wtf
here's a nix-darwin timer: https://codeberg.org/adamcstephens/dotfiles/src/commit/fd44a309d17aab3a363f5ee978922a6365e19830/parts/darwin.nix#L93 So this is probably what I'm looking for; thanks! | 13:44:00 |
| cole-h | $ ssh ... df -h
Filesystem Size Used Avail Capacity iused ifree %iused Mounted on
/dev/disk3s1s1 228Gi 8.9Gi 59Mi 100% 356091 599320 37% /
$ # run the cleanup command
$ df -h
Filesystem Size Used Avail Capacity iused ifree %iused Mounted on
/dev/disk3s1s1 228Gi 8.9Gi 20Gi 31% 356091 208468320 0% /
not the most cleanup I've seen from it before but "lol"
| 13:46:25 |
| Lily Foster | In reply to @cole-h:matrix.org
$ ssh ... df -h
Filesystem Size Used Avail Capacity iused ifree %iused Mounted on
/dev/disk3s1s1 228Gi 8.9Gi 59Mi 100% 356091 599320 37% /
$ # run the cleanup command
$ df -h
Filesystem Size Used Avail Capacity iused ifree %iused Mounted on
/dev/disk3s1s1 228Gi 8.9Gi 20Gi 31% 356091 208468320 0% /
not the most cleanup I've seen from it before but "lol"
wait what's the cleanup command? i didn't know you had to clean up rosetta stuff 🤔 | 13:59:55 |
| cole-h | sudo /System/Library/Filesystems/apfs.fs/Contents/Resources/apfs.util -P -minsize 0 /System/Volumes/Data
| 14:23:50 |
 Artturin | In reply to @cole-h:matrix.org
sudo /System/Library/Filesystems/apfs.fs/Contents/Resources/apfs.util -P -minsize 0 /System/Volumes/Data
https://github.com/search?q=%2FSystem%2FLibrary%2FFilesystems%2Fapfs.fs%2FContents%2FResources%2Fapfs.util+-P+-minsize+0+%2FSystem%2FVolumes%2FData&type=code | 21:21:10 |
| Artturin | In reply to @cole-h:matrix.org
sudo /System/Library/Filesystems/apfs.fs/Contents/Resources/apfs.util -P -minsize 0 /System/Volumes/Data
* https://github.com/search?q=%2FSystem%2FLibrary%2FFilesystems%2Fapfs.fs%2FContents%2FResources%2Fapfs.util+-P+-minsize+0+%2FSystem%2FVolumes%2FData+NOT+is%3Afork&type=code | 21:22:38 |
| 29 Oct 2023 |
 K900 | Interesting failure-adjacent mode here actually | 19:53:00 |
| K900 | https://github.com/NixOS/nixpkgs/pull/263768 broke eval of the package on aarch64 | 19:53:10 |
| K900 | But the check is green, because the other package evaluated and built | 19:53:18 |
| 30 Oct 2023 |
 Hubble the Wolverine (they/them) | I'm getting an interesting failure on https://github.com/NixOS/nixpkgs/pull/264292
error: evaluation aborted with the following error message: 'Failed to evaluate libhugetlbfs-2.23: «unknown-meta»: has an invalid meta attrset:
- key 'meta.badPlatforms' has invalid value; expected list of (string or attribute set of anything), got
is this a known one? I think I saw it in #dev:nixos.org before...
| 04:43:27 |
| Hubble the Wolverine (they/them) | In reply to @hubofeverything:bark.lgbt
I'm getting an interesting failure on https://github.com/NixOS/nixpkgs/pull/264292
error: evaluation aborted with the following error message: 'Failed to evaluate libhugetlbfs-2.23: «unknown-meta»: has an invalid meta attrset:
- key 'meta.badPlatforms' has invalid value; expected list of (string or attribute set of anything), got
is this a known one? I think I saw it in #dev:nixos.org before...
I'm assuming it's related to https://github.com/NixOS/nixpkgs/pull/264290 | 04:49:31 |
| Hubble the Wolverine (they/them) | Oh I need to rebase I think | 04:51:58 |
|  Anthony Rsl joined the room. | 17:04:25 |
| 1 Nov 2023 |
|  cafkafk left the room. | 10:42:47 |
|  Ramses 🇵🇸 changed their display name from Ramses to Ramses 🇵🇸. | 19:58:24 |
| 2 Nov 2023 |
 7c6f434c | Hm, how to interpret «Complete, with errors» without details for ofborg-eval? | 07:23:06 |
| 7c6f434c | ofborg-eval-darwin complains about crashed Nix daemon (ENOENT on a socket) though | 07:23:38 |
| 10 Nov 2023 |
|  Luke joined the room. | 05:58:55 |
| Luke | Is there a way to check where a particular build request from a given PR is in the queue? I don’t know enough about Prometheus to know if this is feasible or to construct this query myself | 06:02:20 |
 hexa | no, that information is not available | 11:12:13 |
| Luke | Okay, in that case are there any timestamps within the logs or ways to correlate builds triggered from a PR to when they happened or what the inputs were? I’m trying to get something building on Darwin and without timestamps I can’t tell which run was the latest? | 14:27:14 |
