| 30 Oct 2026 |
 autra | same for graphopper, seems to be in the caceh. | 16:49:17 |
| autra | ah no, for graphopper I fixed it locally and forgot :laughing: | 16:50:39 |
| 31 Oct 2026 |
 Ivan Mincik (imincik) | Nice work, thank you. We are getting close to zero failures for geospatial on Linux - https://github.com/imincik/nix-utils/actions/runs/41051002900 | 08:56:59 |
| Ivan Mincik (imincik) | * Nice work, thank you. We are getting close to zero failures for geospatial on x86-64 Linux - https://github.com/imincik/nix-utils/actions/runs/41051002900 | 08:57:30 |
| Ivan Mincik (imincik) | Not bad on Darwin as well - https://github.com/imincik/nix-utils/actions/runs/41050439146 | 08:59:28 |
| 1 Nov 2026 |
|  Jassu joined the room. | 14:18:26 |
| 3 Nov 2026 |
| autra | Yes, I'm currently dealing with martin, but some tests failures are just odd. I'm tempting to just exclude them :-) | 09:38:28 |
| 4 Nov 2026 |
|  phanirithvij (loudgolem) changed their display name from phanirithvij/loudgolem to phanirithvij (loudgolem). | 07:18:44 |
| 5 Nov 2026 |
 r-burns | Quick heads up, not sure if you guys are typically involved in staging stabilization, but it looks like the current staging-next iteration has some test failures in geopandas that will be a blocker for some geospatial packages. Might be a good idea to take a look before the staging-next merge. Example: https://hydra.nixos.org/build/352035116/step/1/log | 00:46:30 |
| Ivan Mincik (imincik) | Looks like geopandas tests can't find shapely
try:
import shapely
if Version(shapely.__version__) < Version("2.0.0"):
shapely = None
except ImportError:
shapely = None
if not shapely:
> raise ValueError("'mask' parameter requires Shapely >= 2.0")
E ValueError: 'mask' parameter requires Shapely >= 2.0
| 10:14:15 |
| Ivan Mincik (imincik) | I can't git pull nixpkgs from Github now. | 10:15:27 |
| Ivan Mincik (imincik) | It is hanging forever | 10:15:39 |
| Ivan Mincik (imincik) | Shapely is reporting wrong version. I am testing the fix now. | 11:03:56 |
| Ivan Mincik (imincik) | * r-burns: Shapely is reporting wrong version. I am testing the fix now. | 11:04:11 |
| Ivan Mincik (imincik) | ix-shell -I nixpkgs=. -p 'python3.withPackages (p: [p.shapely])' --run 'python -c "import shapely; print(shapely.__version__)"'
| 11:10:24 |
| Ivan Mincik (imincik) | * nix-shell -I nixpkgs=. -p 'python3.withPackages (p: [p.shapely])' --run 'python -c "import shapely; print(shapely.__version__)"'
| 11:10:28 |
| Ivan Mincik (imincik) | r-burns: which branch I should target with my PR ? | 11:21:09 |
| Ivan Mincik (imincik) | * r-burns: which branch I should target with my PR - staging ? | 11:21:30 |
| Ivan Mincik (imincik) | https://github.com/NixOS/nixpkgs/pull/583583 | 11:35:21 |
| 6 Nov 2026 |
| r-burns | I think you should target staging-next since it's a fixup for the current staging-next cycle. Looks good, I'll repost it for the staging folks | 02:16:29 |
| 8 Nov 2026 |
 sikmir | https://github.com/NixOS/nixpkgs/pull/584377 | 14:23:08 |
| 9 Nov 2026 |
| Ivan Mincik (imincik) | Packages failing on Linux - https://github.com/imincik/nix-utils/actions/runs/42480206786 | 07:41:13 |
| Ivan Mincik (imincik) | Faling on Darwin - https://github.com/imincik/nix-utils/actions/runs/42479732708 | 07:41:39 |
| 10 Nov 2026 |
| autra | This is ready again (fixing graphhopper build) : https://github.com/NixOS/nixpkgs/pull/581051 | 09:01:55 |
| autra | the hash will change again for a variety of reason, but my understanding is that it's unavoidable with nix+mvn | 09:02:23 |
| autra | at least it may stay stable enough for graphhopper builds to go into the cache | 09:02:43 |
| autra | and then, it would matter only for updates (but also for mvn + plugins update, and any dependencies unfortunately) | 09:03:09 |
| autra | I'm still dealing with martin. | 09:03:26 |
| autra | we do have tough build error these days... | 09:04:39 |
|  imincik joined the room. | 11:23:37 |
