| 25 May 2021 |
 jonringer | there's a fix in both staging-next's for it | 18:29:45 |
| jonringer | but the initial bump caused some regressions | 18:29:57 |
 Jan Tojnar | does anyone remember the tool that scanned configure.ac and edited package expression with missing dependencies? | 18:31:04 |
 ris_ | ahyes that looks like https://gitlab.gnome.org/GNOME/libxml2/-/issues/255 maybe? | 18:31:06 |
 hexa | ah, might need to roll back https://github.com/NixOS/nixpkgs/pull/124320 then | 18:34:18 |
| hexa | if we get this fixed in libxml2 | 18:34:28 |
| jonringer | if you build it on staging-next, you should get a better result, post-patch | 18:39:04 |
 lukegb (he/him) | Hmm, is Darwin's stdenv broken on master | 20:53:00 |
| lukegb (he/him) | https://hydra.nixos.org/build/144116748 | 20:53:03 |
 dotlambda | In reply to @hexa:lossy.network
ah, might need to roll back https://github.com/NixOS/nixpkgs/pull/124320 then You mean https://github.com/NixOS/nixpkgs/pull/124385? | 21:07:22 |
| hexa | dotlambda: we shouldn't disable the faulty tests on either package if we have fix in staging-next, or at least roll them back after the fix has been merged, as the pywemo disabledTests already landed in master. | 21:08:37 |
| jonringer | I'm fine with disabling on master/release-21.05, but we should revert the patch on staging-next, as it's no longer needed | 21:09:35 |
| hexa | how far along is staging-next/staging-21.05-next? | 21:10:44 |
| hexa | as in: is it worth it time-wise? | 21:11:07 |
| hexa | and does the latter contain said fix? | 21:11:23 |
| jonringer | both contain the libxml2 patch | 21:17:45 |
| jonringer | and they probably won't be getting merged until friday at the earliest | 21:18:05 |
| jonringer | Hydra is being crushed under its own weight https://hydra.nixos.org/queue_summary | 21:18:25 |
| hexa | tbf, home-assistant isn't really affected, unless someone uses the wemo component, since it's only part of the check inputs | 21:18:52 |
| hexa | so I don't worry much | 21:18:57 |
| hexa | friday sgtm | 21:19:00 |
| jonringer | like I said, just doing the revert patch on the staging-next* branches is also fine with me. And will probably help prevent people from looking into fixing it for ZHF as well | 21:19:39 |
| hexa | In reply to @hexa:lossy.network
tbf, home-assistant isn't really affected, unless someone uses the wemo component, since it's only part of the check inputs for context: we don't run the tests for endusers by default, because they take quite some time to run | 21:19:49 |
| hexa | sgtm, then we merge https://github.com/NixOS/nixpkgs/pull/124395 | 21:20:17 |
| jonringer | oh, I know
=========================== short test summary info ============================
FAILED tests/components/prometheus/test_init.py::test_view[pyloop] - assert '...
= 1 failed, 11713 passed, 8 skipped, 2 xfailed, 7816 warnings in 1084.71s (0:18:04) =
DEBUG:pysonos.data_structures_entry:pysonos.data_structures_entry imported
DEBUG:pyvera:DEBUG logging is
| 21:20:36 |
| hexa | right, and then there is this flaky humidity test | 21:20:58 |
| jonringer | yes, >:( we are well acquainted | 21:21:18 |
| hexa | too well | 21:21:29 |
| jonringer | I reviewed 4 python package pr's all of which triggered, home-assistant. 3 failed, and 1 passed | 21:21:56 |
| hexa | and it compares the output of the whole prometheus exporter page, and doesn't show me the full vars, so hard to say what goes wrong exactly | 21:21:57 |
