| 10 Oct 2025 |
 dramforever | presumably someone is testing them before release... | 07:51:55 |
 Grimmauld (any/all) | In reply to @dramforever:matrix.org
Grimmauld (any/all): python3Packages.inkex
ERROR inkex/tester/test_inx_file.py - lxml.etree.SchematronError: lxml.etree was compiled without Schematron supp...
ERROR tests/test_inkex_inx.py - lxml.etree.SchematronError: lxml.etree was compiled without Schematron supp...
https://gitlab.gnome.org/GNOME/libxml2/-/releases/v2.15.0
Support for Schematron is now disabled by default.
Uh oh.... | 08:14:34 |
| Grimmauld (any/all) | Worst case you could technically maybe pin libxml2_13 which we do already have, but I'd rather not... I can draft an override flag (something like lxml.override { schema tonsil port = true;} that gets passed through to libxml, would make the overrides more manageable maybe | 08:17:17 |
| Grimmauld (any/all) | If its test-only, then skip makes sense. But I am worried there may be more of these hidden in the tree. Testing libxml updates is hard, I'll add this to the list of things I'll test next time | 08:19:06 |
 vcunat | Maybe notify upstream about schematron deprecation? | 08:29:03 |
 K900 | Since I feel like people have asked themselves this | 08:31:09 |
| K900 | https://en.wikipedia.org/wiki/Schematron | 08:31:11 |
| K900 | It's like XSLT but different | 08:31:16 |
| dramforever | yeah that's the thing i'm not sure about, i've been checking but also have no idea how | 08:33:54 |
| dramforever | the best part of course is this is some "pre-ISO-Schematron" thing | 08:34:15 |
| dramforever | which, i guess it means it's like XSLT but different, but more different | 08:34:43 |
| dramforever | 2.15 still has --with-schematron at least, so at least we can keep overriding things | 08:36:28 |
| K900 | Just kill it | 08:37:09 |
| dramforever | just in case, i meant... hopefully nothing critical is really depending on having schematron | 08:38:33 |
| Grimmauld (any/all) | Tbh if there is not more that breaks, we just kill it. If there is significant packages that have a real dependency, I can draft an override flag. | 08:46:27 |
| dramforever | "kill it" as in "do nothing"? | 08:47:36 |
| Grimmauld (any/all) | In reply to @dramforever:matrix.org
"kill it" as in "do nothing"? Yes | 08:48:26 |
| K900 | Kill it as in skip the tests | 08:48:28 |
| K900 | And ignore it | 08:48:31 |
| dramforever | "kill it" -> https://github.com/NixOS/nixpkgs/pull/450599 | 08:48:50 |
| dramforever | and pray that's it | 08:49:06 |
| dramforever | * and pray that that's it | 08:49:08 |
| Grimmauld (any/all) | Well yes of course we do have to actively fix the soft dependencies (like test fails and optional features) depending on libxml schematron | 08:49:10 |
 emily | sorry for being absent the entire -next 🫠| 16:31:18 |
| emily | will probably still have reduced activity for a bit. hope LLVM 21 went okay. I see CMake 4 has been exciting. | 16:31:43 |
 Yureka (she/her) | firefox wasm stuff is broken on staging | 18:07:05 |
| Yureka (she/her) | specifically wasilibc-static-wasm32-unknown-wasi-27 | 18:08:13 |
| Yureka (she/her) | (not on staging-next though) | 18:08:36 |
|  hougo joined the room. | 18:25:07 |
 Lun | might be related to https://github.com/NixOS/nixpkgs/pull/442945 ? seems to be upset that _LIBCPP_HARDENING_MODE is defined | 19:40:55 |
