| 29 May 2026 |
| @koichiio:matrix.org joined the room. | 08:21:23 |
Matt Sturgeon | In reply to @llakala:matrix.org is there any reason to prefer isNull over x == null? Wasn't isNull soft-deprecated at one point, and the deprecation was removed because valid use cases were identified? Maybe I'm misremembering.
Generally speaking, == null and != null is simpler and more readable, and avoids invoking a function.
Same for ? vs hasAttr imo, although I don't think that one was ever deprecated. | 08:43:01 |
hustlerone | anything to compact my code will do | 08:51:13 |
Matt Sturgeon | nix-repl> options.services.pixelfed.nginx.type.description
"null or (submodule)"
nix-repl> options.services.pixelfed.nginx.valueMeta
{ }
Looks like an oversight in types.nullOr, not propagating its elem-type's valueMeta. In fact, nullOr doesn't seem to implement v2 check+merge at all (nullOr's merge fn).
BTW, this room is really meant for discussing the nix language itself. For the module system there is https://matrix.to/#/#modules:nixos.org
| 08:52:51 |
toonn | OK, I'll head over there. | 08:55:35 |
| bigbird99 joined the room. | 22:33:13 |
| 30 May 2026 |
| qweru changed their profile picture. | 10:28:06 |
| @lumi:the-apothecary.club removed their profile picture. | 17:58:42 |
| sbstratos79 joined the room. | 20:23:38 |
| 31 May 2026 |
| Aliaksandr set a profile picture. | 00:31:03 |
| Schmiddi joined the room. | 10:31:33 |
| @lumi:the-apothecary.club set a profile picture. | 12:34:09 |
| @lumi:the-apothecary.club changed their profile picture. | 12:50:47 |
| @lumi:the-apothecary.club changed their profile picture. | 15:29:25 |
| @lumi:the-apothecary.club changed their profile picture. | 15:34:35 |
| @lumi:the-apothecary.club changed their profile picture. | 17:22:20 |
| @lumi:the-apothecary.club changed their profile picture. | 18:28:26 |
| Vandy Carlos changed their display name from vandycarlos to Vandy arlos. | 22:00:30 |
| Vandy Carlos changed their display name from Vandy arlos to Vandy Carlos. | 22:00:39 |
| 1 Jun 2026 |
| @lumi:the-apothecary.club changed their profile picture. | 14:19:30 |
toonn | There's no built-in way to detect whether a value is recursive and if it is at what points it recurses, right? | 15:16:14 |
| Tanja (she/her) 📞 TNJA (8652) changed their display name from Tanja (she/her) to Tanja (she/her) 📞 TNJA (8652). | 15:33:24 |
alexfmpe | That sounds undecidable | 15:46:02 |
alexfmpe | maybe semi-undecidable if you trace calls | 15:46:46 |
| @lumi:the-apothecary.club changed their profile picture. | 15:53:09 |
| @lumi:the-apothecary.club changed their profile picture. | 15:57:23 |
| @lumi:the-apothecary.club changed their display name from Orion to Vel. | 15:59:31 |
toonn | I wouldn't be surprised if attrsets boil down to pointers and checking whether a pointer occurs in the value it points to is not undecidable. | 16:28:54 |
Qyriad | Unfortunately: thunks | 16:30:28 |
Qyriad | Attrsets do essentially boil down to pointers, but the expressions they point to are essentially virtual function calls | 16:31:40 |