| 21 Mar 2026 |
 raitobezarius | no one knows much about matrix, don't worry | 22:11:05 |
| 22 Mar 2026 |
|  @n0toose:ccc.ac left the room. | 12:39:35 |
|  KFears& 🏳️⚧️ (they/them) changed their display name from KFears 🏳️⚧️ (they/them) to KFears& 🏳️⚧️ (they/them). | 13:19:59 |
|  kungpaochicken joined the room. | 17:07:15 |
 Liyua | Is there a way to make nix output monitor show waiting tasks as waiting instead of just showing them as continuing to build?
| 20:12:34 |
| Liyua | With normal nix only the active tasks are being shown as being built right now and the waiting ones are marked as waiting (and info on how many others they're waiting)
| 20:13:49 |
| Liyua | With lix all of them show up as "building" | 20:14:46 |
 whispers [& it/fae] | i think you're seeing the opposite bug: nom does not correctly determine when builds have finished with lix | 20:15:01 |
| whispers [& it/fae] | it's a known bug | 20:15:17 |
| Liyua | It's mostly just a vanity thing but it still bothers me for some reason | 20:15:23 |
| Liyua | That might also be the case | 20:15:36 |
| whispers [& it/fae] | there's been some attempts to try to fix it, but last i checked, it requires some lix-side changes to do correctly. there have been many previous discussions in #nix-output-monitor:nixos.org | 20:16:05 |
| whispers [& it/fae] | * there's been some attempts to try to fix it, but last i checked, it requires some lix-side changes to do correctly (i.e. none of the attempts have been completely successful). there have been a few previous discussions in #nix-output-monitor:nixos.org | 20:16:29 |
| Liyua | The thing I'm seeing is an 8 minute build time for tmux.conf because tmux.conf depends on nvim and nvim depends on nvim.conf and nvim.conf has to build some plugins from source
| 20:16:30 |
| whispers [& it/fae] | Redacted or Malformed Event | 20:17:35 |
| whispers [& it/fae] | * yeah, ime e.g. tmux.conf is done building in a few seconds, nom just doesn't realize that | 20:17:51 |
| Liyua | With nix it'd just show as "waiting" I think until all dependencies are done building and the hash for the store paths are known | 20:18:21 |
| Liyua | It's just a very minor bother tho so if there's not a solution like "hey just set these options in your config" I don't think it's worth it for me to keep looking. Still thanks for taking your time to respond <3
| 20:19:42 |
|  Enrico Cinti joined the room. | 22:39:15 |
| 23 Mar 2026 |
|  ShalokShalom changed their profile picture. | 02:05:21 |
|  nightuser changed their profile picture. | 11:30:45 |
 iggle | Since I can't find any mentions of it in this chat, nixpkgs or lix issues: Lix 1.94 currently doesn't build for the nixpkgs-darwin-25.11 channel due to python3.13-pyxattr not being available on aarch64-darwin yet. I see that @raitobezarius:matrix.org merged a PR into master 2 months ago to enable this platform, however this hasn't reached the 25.11 channel yet somehow. This is not an issue if you get lix from the normal nixos-25.11 channel, as that lix derivation didn't yet add pyxattr to its buildInputs. However, somehow the pyxattr buildInput snuck into nixpkgs-25.11-darwin. I have no idea how this channel is coordinated though. | 13:42:58 |
 hexa | Backporting https://github.com/NixOS/nixpkgs/pull/476848 | 13:57:48 |
| raitobezarius | Let's backport it | 14:02:50 |
 llakala | this cl made me realize a fun nix party trick | 15:17:07 |
| llakala | realizing that nix allowed empty strings as a key made me wonder whether you could actually access a variable of that name | 15:17:45 |
| llakala | let
"" = 5;
in null
| 15:17:52 |
| llakala | how would one access this? | 15:17:57 |
| llakala | of course you can access it when its in an attrset | 15:18:12 |
| llakala | let
x."" = 5;
in x."" # 5
| 15:18:28 |
