Conversation
vcunat
commented
Jul 21, 2022
- workflow docs: nixos.org/manual/nixpkgs/unstable/#submitting-changes-commit-policy
- important job(s): hydra.nixos.org/job/nixpkgs/staging-next/unstable#tabs-constituents
- jobset: hydra.nixos.org/jobset/nixpkgs/staging-next
- nix-review reports: https://malob.github.io/nix-review-tools-reports/
- previous staging-next: staging-next 2022-07-11 #181133
|
The e-d-s failure looks like something caused by the recent g-i fiddling, cc @Artturin |
| inherit version; | ||
| sha256 = "sha256-ConK3ABipeU2ZN3gQ/bAlxcrjBxfAJRJAJUoL/mZWl8="; | ||
| }; | ||
| }); |
There was a problem hiding this comment.
We don't override packages inside python-modules, because it is easily results in conflicts when multiple packages depend on different versions while only one can claim the name inside the python environment.
If you want to fix matrix-synapse OTOH, an override there would be acceptable.
There was a problem hiding this comment.
This is only used in tests (and in checkInputs), so I think it's fine here?
There was a problem hiding this comment.
Ah well, if it's not propagated that should be fine.
There was a problem hiding this comment.
It is not, and it doesn't look like anything in nixpkgs uses both pysaml2 and pymongo at the same time.
There was a problem hiding this comment.
it doesn't look like anything in nixpkgs uses both pysaml2 and pymongo at the same time.
but something could use it in the future. Don't we have a hook now that detects it? But I don't think it is perfect.
As long as this never gets propagated it should be fine ™️
There was a problem hiding this comment.
Something can even get away with using both, as long as it doesn't specifically use pymongo and the mongo storage for pysaml2, but if you're doing that, you'll explode anyway.
on it... look's like libical isn't building its gir and vapi files on master EDIT its because libical needs to be explicitly told to build them |
the build doesn't autodetect the existence of the programs required
Security fixes
Let's stop using src.override. I see no advantage, and the switch of original src from `sha256 =` to `hash =` was causing really confusing errors.
|
The impact of the python package updates is not small, but I don't think that delaying merge would help significantly. (discussed with others on https://matrix.to/#/#staging:nixos.org ) |