25 Jun 2025 |
maralorn | We totally could. But I like them too much to do it myself. 😂 | 16:46:56 |
maralorn | What the heck is the correct way to run doctests in a nixpkgs package? | 21:20:33 |
maralorn | cabal repl --with-compiler=doctest does not work because the doctest does not have the same pkg db than the ghc used by cabal. | 21:21:10 |
sterni | cabal exec -- cabal repl --with-compiler=doctest maybes | 22:16:44 |
sterni | * cabal exec -- cabal repl --with-compiler=doctest maybe? | 22:16:46 |
maralorn | Still the same problem:
Configuring library for status-script-0.1.0.0...
Loaded package environment from /home/maralorn/git/config/packages/status-script/dist-newstyle/tmp/environment.-175358/.ghc.environment.x86_64-linux-9.8.4
<command line>: cannot satisfy -package-id aeson-2.2.3.0-2mnLs4rnok4A0ksXaTX6dm
and I am relatively certain that is because cabal-install does use a different ghc pkg db.
| 22:34:42 |
maralorn | * Still the same problem:
Configuring library for status-script-0.1.0.0...
Loaded package environment from /home/maralorn/git/config/packages/status-script/dist-newstyle/tmp/environment.-175358/.ghc.environment.x86_64-linux-9.8.4
<command line>: cannot satisfy -package-id aeson-2.2.3.0-2mnLs4rnok4A0ksXaTX6dm
and I am relatively certain that is because cabal-install does use a different ghc pkg db than doctest.
| 22:34:46 |
maralorn | So this seems to me that we should maybe patch doctest so that it picks up the package db from the ghc.withPackages it is used from? | 22:35:25 |
26 Jun 2025 |
sterni | ah yes I remember https://github.com/haskell/cabal/issues/7792#issuecomment-2124304244 | 21:25:04 |
28 Jun 2025 |
Peter Becich | On haskell-updates branch, ./maintainers/scripts/haskell/regenerate-hackage-packages.sh produces
...
hackage2nix: user error (cannot parse cabal file /nix/store/kmz1d73l5k1nhv5269dl5i68yjwlbcir-unpacked-cabal-hashes/pms-application-service/0.0.3.0/pms-application-service.cabal)
hackage2nix: thread blocked indefinitely in an MVar operation
Does anyone else see this or is it local to me?
| 20:22:52 |
29 Jun 2025 |
chreekat | In reply to @peterbecich:matrix.org
On haskell-updates branch, ./maintainers/scripts/haskell/regenerate-hackage-packages.sh produces
...
hackage2nix: user error (cannot parse cabal file /nix/store/kmz1d73l5k1nhv5269dl5i68yjwlbcir-unpacked-cabal-hashes/pms-application-service/0.0.3.0/pms-application-service.cabal)
hackage2nix: thread blocked indefinitely in an MVar operation
Does anyone else see this or is it local to me?
I think the real error is getting swallowed somewhere. I know from fixing Stackage recently that that package uses a newer version of Cabal file format, so hackage2nix (or whatever underlying tool) probably also needs to be built with an updated version of the cabal library | 07:11:34 |
levin | I'm working on a pull-request right now (https://github.com/NixOS/nixpkgs/pull/419877). In the last unresolved comment, I'm asked to get the tests for the package (haskore), which the pull-request originally disabled using dontCheck , working.
For these tests to work, I would need to have some files (https://archives.haskell.org/code.haskell.org/haskore/revised/core/src/Test/) in the src, which aren't there because they aren't listed in the cabal file under extra-source-files .
How would I resolve this? I couldn't really find anything...
| 07:41:19 |
maralorn | I think that is more a factual correction than a demand. | 09:00:26 |
ners | Would it be possible to cache the pkgsStatic GHC for aarch64 as well, in addition to x86_64 linux? | 16:26:11 |
ners | In Hydra, that is. Currently we can get the musl GHC from haskell-updates for x86_64, but not for aarch64. | 16:26:45 |
ners | * In Hydra, that is. Currently we can get the cached musl GHC from haskell-updates for x86_64, but not for aarch64. | 16:26:53 |
Tom Westerhout | Following up on ners's point, there's this comment in release-haskell.nix : https://github.com/NixOS/nixpkgs/blob/dcc4da22e1bf5ed04220e7c5ddaa80abc0dee646/pkgs/top-level/release-haskell.nix#L409
# pkgsMusl is compiled natively with musl. It is not
# cross-compiled (unlike pkgsStatic). We can only
# natively bootstrap GHC with musl on x86_64-linux because
# upstream doesn't provide a musl bindist for aarch64.
I'm wondering if it still stands because I was able to build stuff locally
| 16:27:44 |
Tom Westerhout | Ah wait, never mind. I can't read. I was reading pkgsMusl instead of pkgsStatic . For pkgsStatic , it says "times out on Hydra" | 16:29:21 |
Tom Westerhout | * Ah wait, never mind. I can't read. I was reading pkgsMusl instead of pkgsStatic . For pkgsStatic , it says "times out on Hydra". Is there any chance to get an exception there? | 16:29:49 |
Alex | In reply to @ners:nixos.dev Would it be possible to cache the pkgsStatic GHC for aarch64 as well, in addition to x86_64 linux? My guess, without looking at Hydra, is that it could potentially break the path size limit. | 17:17:56 |
Peter Becich | this fixes it, rebuilding the tool as you suggested, review please: https://github.com/NixOS/nixpkgs/pull/421127 | 21:58:21 |
30 Jun 2025 |
ners | maralorn any thoughts? | 09:51:30 |
maralorn | I am sadly quite unfamiliar with the policys around hydra timeouts. I guess there have to exist exceptions for something like chromium? But I think in Haskell world we never had one? | 10:06:57 |
maralorn | * I am sadly quite unfamiliar with the policies around hydra timeouts. I guess there have to exist exceptions for something like chromium? But I think in Haskell world we never had one? | 10:07:10 |
1 Jul 2025 |
sterni | ners: in principle the timeout is configurable per derivation. That comment is from 2021, though, so it may no longer be accurate for the builders we currently have. Currently, it doesn't make a difference since haskell-updates is not tested on aarch64-linux at all. I don't know when/if that is going to change. Also, haskell-updates is relatively disconnected from master these days, so you will get cache misses even on x86_64-linux (haskell-updates is for catching regressions primarily, not populating the binary cache). Feel free to open a PR adding jobs for pkgsStatic haskell to release.nix or maybe release-cross.nix which are intended for testing/populating binary cache on master. | 09:54:04 |
sterni | going to merge h-u this evening probably | 16:01:25 |
sterni | staging-next cycle is probably starting soon | 16:01:36 |
maralorn | @sternenseemann:systemli.org At this point I have the feeling our rotation is kinda broken. | 16:43:52 |
maralorn | As in theoretically it's my two weeks to do the merging. | 16:44:47 |
maralorn | I mean I am not complaining. | 16:45:06 |