Sender | Message | Time |
---|---|---|
26 Jul 2024 | ||
For others checkin in here, gonna work around it: https://github.com/NixOS/nixpkgs-check-by-name/pull/90 | 23:44:05 | |
Philip Taron (UTC-8): Oh no, the Cargo.lock is still at 0.1.1, why did this not break anything 😆 | 23:46:07 | |
Must be some consequence of the cargo update changes?? | 23:48:10 | |
I noticed it from those: https://github.com/NixOS/nixpkgs-check-by-name/actions/runs/10118954269#summary-27986608230 | 23:48:51 | |
But like, I'd expect the build itself to fail from this | 23:49:09 | |
Not sure what the consequences of a mismatched Cargo.{toml,lock} are | 23:49:47 | |
Philip Taron (UTC-8): I guess I should finish the Changelog PR, because there it would be handled automatically: https://github.com/NixOS/nixpkgs-check-by-name/pull/46/files#diff-d9002f82f23cc9b3c1545ec7d5ddfc3911aec8ba24d4e3d21c73f40355da2a16R85
This updates both | 23:55:42 | |
27 Jul 2024 | ||
* Philip Taron (UTC-8): I guess I should finish the Changelog PR, because there it would be handled automatically: https://github.com/NixOS/nixpkgs-check-by-name/pull/46/files#diff-d9002f82f23cc9b3c1545ec7d5ddfc3911aec8ba24d4e3d21c73f40355da2a16R85
This updates the version in both | 00:05:49 | |
29 Jul 2024 | ||
10:01:47 | ||
1 Aug 2024 | ||
11:40:49 | ||
3 Aug 2024 | ||
14:24:05 | ||
Let me know if I should move this to #dev:nixos.org, but I have a few questions about I’m working on the Darwin SDK refactor. As part of that work, I want to move the SDK and source-releases out of For the SDK, there will be multiple versions. For how to handle those, I can override
For the source-releases, I need a scope. The SDK overrides packages in the scope to the versions it needs for that SDK version (e.g., the 11.3 SDK uses Csu 85 while the 14.4 SDK uses Csu 88). I know things in We’ve been talking about various options in #macos:nixos.org. One possibility is keeping it as a package with the other packages not nested (so | 14:24:14 | |
(FWIW my proposal was to do by-name/ap/apple-oss-* but have an apple-oss scope defined outside of by-name that just does { Csu = apple-oss-Csu; … } so that the various packages can still depend on apple-oss and get the scope overrides atomically – not sure if that's what Randy meant by "keeping it as a package with the other packages not nested") | 14:25:59 | |
I do think that it's just wrong for the checker to complain about foo = callPackage ./bar { } if the result of foo is a scope or attrset or something instead of a derivation. (I don't know if it actually does or not, but Randy was worried about it) | 14:26:47 | |
unless the intent is that nobody should ever add a top-level scope ever again | 14:26:56 | |
I was assuming a by-name/ap/apple-oss that provided a scope, but the packages were packaged separately. I wouldn’t be able to reference a scope defined outside of the SDK. | 14:28:10 | |
I guess I could define the scope in the SDK. That seems like a lot of boilerplate though. | 14:28:49 | |
I was thinking apple-oss = callPackage ./os-specific/darwin/apple-oss { }; that just provides the scope and re-exports the top-level stuff from pkgs/by-name ? | 14:29:10 | |
which still gets you all the actual new packages being in by-name | 14:29:19 | |
6 Aug 2024 | ||
In reply to @emilazy:matrix.orgIt complains if you return anything other than a derivation while putting something in pkgs/by-name : https://github.com/NixOS/nixpkgs-check-by-name/blob/main/src/eval.rs#L316-L321 | 17:36:34 | |
Here's the branch which controls allowing non-by-name expressions which don't resolve to derivations: https://github.com/NixOS/nixpkgs-check-by-name/blob/main/src/eval.rs#L594-L596 | 17:47:46 | |
8 Aug 2024 | ||
18:48:35 | ||
13 Aug 2024 | ||
03:39:45 | ||
14 Aug 2024 | ||
13:25:53 | ||
19 Aug 2024 | ||
08:51:41 | ||
20 Aug 2024 | ||
12:44:13 | ||
23 Aug 2024 | ||
No office hour today unless somebody else wants to join ;) | 17:57:48 | |
18:25:24 | ||
19:00:49 | ||
24 Aug 2024 | ||
19:43:03 |