| 31 Jan 2024 |
 Shyim | that libxml problem seems also with PHP 8.1 from nixpkgs :D | 14:20:08 |
| Shyim | * that libxml problem seems also with PHP 8.1 from nixpkgs :D, not only nix-phps problem | 14:20:18 |
 Pol | Really ? | 17:16:53 |
| Pol | Care to elaborate? | 17:16:57 |
| Pol | We are close to something here: https://github.com/fossar/nix-phps/actions/runs/7732636096/job/21082991336 | 22:05:43 |
| Pol | Heading to bed. I'm almost done with libxml and PHP stuff, only 1 failure. | 22:34:49 |
| Pol | I'll continue asap | 22:34:59 |
| 1 Feb 2024 |
 Jan Tojnar | In reply to @drupol:matrix.org
We are close to something here: https://github.com/fossar/nix-phps/actions/runs/7732636096/job/21082991336 That requires cast like here https://gitlab.com/libvirt/libvirt-glib/-/merge_requests/53/diffs#b684a5b6959188e3400acdcf5c7196b92fab9fa6_201_204 | 07:25:55 |
| Jan Tojnar | Darwin uses clang compiler which usually increases strictness (e.g. promotes warnings to errors) and then gcc follows a cycle later | 07:27:52 |
| Jan Tojnar | * Darwin uses Clang compiler which usually increases strictness (e.g. promotes warnings to errors), and then gcc follows a cycle later | 07:28:15 |
| Jan Tojnar | * Darwin uses Clang compiler, which usually increases strictness (e.g. promotes warnings to errors), and then gcc follows a cycle later | 07:28:32 |
| Jan Tojnar | * Darwin uses Clang compiler, which often increases strictness in major releases (e.g. promotes warnings to errors), and then gcc follows a cycle later | 07:29:25 |
| Pol | In reply to @jtojnar:matrix.org
That requires cast like here https://gitlab.com/libvirt/libvirt-glib/-/merge_requests/53/diffs#b684a5b6959188e3400acdcf5c7196b92fab9fa6_201_204 There's already a patch being applied for that, I don't understand why it's failing yet. #investigating | 07:42:12 |
| Pol | Jan Tojnar: Could you approve https://github.com/fossar/nix-phps/pull/324 and https://github.com/fossar/nix-phps/pull/323 ? | 07:50:09 |
| Pol | Jan Tojnar: https://github.blog/changelog/2024-01-30-github-actions-macos-14-sonoma-is-now-available/ | 08:20:28 |
| Pol | There's an issue with PHP 8.1 on darwin. The libxml patch is applied here: https://github.com/fossar/nix-phps/pull/321/files#diff-7d838ade1f0467c8bab1cdc2fc34fd867b20eb61cf6e26f50599dde42d776f27R48 but in the build log (https://pipelinesghubeus23.actions.githubusercontent.com/kOCT2mqNiaOhTutxCPe8hYXDv6z0C4UWNeYg9Obg2gmPygfshU/_apis/pipelines/1/runs/2590/signedlogcontent/18?urlExpires=2024-02-01T08%3A56%3A58.9001059Z&urlSigningMethod=HMACV1&urlSignature=OLmIQFW%2F0eJgM%2FfYIoALRA43HCG0X5tttWkPjpG3LHo%3D), I don't see it applied anywhere. Do you have a clue? | 08:59:12 |
| Pol | Someone with a darwin machine is free to help in here :) | 08:59:22 |
| Pol | I think I discovered an issue in phps.nix | 09:04:56 |
| Pol | 
Download image.png | 09:06:13 |
| Pol | Running nix build .#php81 -L doesn't throw the error. | 09:06:32 |
| Pol | I see, that is only applied to PHP versions we maintain in fossar/nix-phps, not in PHP versions that currently are in nixpkgs. | 09:10:53 |
| Pol | That's confusing. | 09:11:07 |
| Pol | Priority PR to merge to fix all of this shit: https://github.com/NixOS/nixpkgs/pull/283175#pullrequestreview-1855951367 | 09:15:27 |
| Pol | Jan Tojnar: We also need this new PR https://github.com/NixOS/nixpkgs/pull/285511 | 09:38:01 |
| Pol | Jan Tojnar: shouldn't we do this instead (instwad of the patch in https://github.com/NixOS/nixpkgs/pull/283175) ? https://github.com/NixOS/nixpkgs/pull/285542 | 13:15:40 |
| Pol | * Jan Tojnar: shouldn't we do this instead (instead of the patch in https://github.com/NixOS/nixpkgs/pull/283175) ? https://github.com/NixOS/nixpkgs/pull/285542 | 13:15:47 |
| Pol | Since the issue in only with Darwin AND php81 | 13:18:18 |
| Pol | I'm entering in a meeting, I'll check your replies later. | 13:18:27 |
| Jan Tojnar | In reply to @drupol:matrix.org
Jan Tojnar: shouldn't we do this instead (instead of the patch in https://github.com/NixOS/nixpkgs/pull/283175) ? https://github.com/NixOS/nixpkgs/pull/285542 That is a hacky workaround that might mask actual issues (especially when we add more patches). Patch is a more proper solution. | 13:20:24 |
| Jan Tojnar | And it is issue with Darwin just for now. It will be an issue everywhere once we bump gcc. | 13:21:13 |
