| 11 Nov 2024 |
 sterni (he/him) | In reply to @manuelbaerenz:matrix.org
is shellFor expected to work? Right now I'm getting "error: no C compiler provided for this platform" shellFor itself works the same way, must have been a similar problem to jsaddle-warp in your example now | 12:12:27 |
| sterni (he/him) | seems from the flags you need to do this
diff --git a/pkgs/development/haskell-modules/configuration-ghcjs-9.x.nix b/pkgs/development/haskell-modules/configuration-ghcjs-9.x.nix
index 448f04228bec..ddba66ec6f5a 100644
--- a/pkgs/development/haskell-modules/configuration-ghcjs-9.x.nix
+++ b/pkgs/development/haskell-modules/configuration-ghcjs-9.x.nix
@@ -20,6 +20,22 @@ with haskellLib;
ghcjs-dom-javascript = addBuildDepend self.ghcjs-base super.ghcjs-dom-javascript;
jsaddle = addBuildDepend self.ghcjs-base super.jsaddle;
jsaddle-dom = addBuildDepend self.ghcjs-base super.jsaddle-dom;
+ jsaddle-warp = super.jsaddle-warp.override {
+ aeson = null;
+ bytestring = null;
+ containers = null;
+ foreign-store = null;
+ http-types = null;
+ jsaddle = null;
+ stm = null;
+ text = null;
+ time = null;
+ transformers = null;
+ wai = null;
+ wai-websockets = null;
+ warp = null;
+ websockets = null;
+ };
entropy = addBuildDepend self.ghcjs-dom super.entropy;
| 12:23:39 |
 Divam | The jsaddle-warp itself should not be built on ghcjs. perhaps jsaddle-warp = null | 12:34:02 |
| sterni (he/him) | jsaddle-warp does appear to be buildable on js though | 13:27:59 |
 Manuel Bärenz | Someone might in theory want to run jsaddle-warp on nodejs maybe 😅 with GHC 9.10 it's not entirely out ouf the question, it should be able to use emscripten to compile the C stuff | 13:46:49 |
| sterni (he/him) | Manuel Bärenz: not really, jsaddle-warp stops pulling in any of warp with javascript https://github.com/ghcjs/jsaddle/blob/b72b4d01fdc186d2c7447beb7273ad13358ff593/jsaddle-warp/jsaddle-warp.cabal#L59-L78 | 15:32:02 |
| 12 Nov 2024 |
 alexfmpe | sterni: can't we grab the haskell-updates parent of the merge with staging, and merge that with staging-next ? | 02:03:34 |
| alexfmpe | then force push | 02:03:41 |
| alexfmpe | * then force push to haskell-updates | 02:03:52 |
|  @ngn999:matrix.org left the room. | 06:59:19 |
 maralorn | branchprotection | 07:44:17 |
| sterni (he/him) | yeah we can't force push anymore unfortunately, otherwise I'd immediately fixed it | 12:09:20 |
| alexfmpe | aaah | 14:20:09 |
| alexfmpe | well, we could close the PR and make a new one? | 14:20:33 |
| alexfmpe | instead of force pushing | 14:21:05 |
| alexfmpe | ah nvm, it's haskell-updates that has the protection, not the PR | 14:21:32 |
| maralorn | Maybe we can delete the branch and create it new? | 14:21:45 |
| alexfmpe | we'd need a totally-not-haskell-updates as well | 14:21:49 |
| maralorn | hurr-durr-i-am-haskell-updates-with-a-mustache | 14:22:42 |
 fgaz | In reply to @sternenseemann:systemli.org
once the package builds you just need to figure out how Cabal installs the js stuff now, they've changed that a bit since ghcjs, i.e. it just emits a jsexe by default and I don't know where the js files end up you'd use in the browser etc. Unfortunately at the moment it doesn't https://git.sr.ht/~fgaz/pointfree-ghcjs/tree/7d871ecd499d1a936132028dc4f34a639a20504f/item/package.nix#L5-8
I should probably open an issue about that... | 14:31:44 |
| sterni (he/him) | alexfmpe: don’t think there’s any urgency we will probably have to wait for staging-next anyways | 14:34:42 |
| sterni (he/him) | contributions can be based against a merge-base of master and haskell-updates for testing | 14:35:10 |
| sterni (he/him) | In reply to @fgaz:matrix.org
Unfortunately at the moment it doesn't https://git.sr.ht/~fgaz/pointfree-ghcjs/tree/7d871ecd499d1a936132028dc4f34a639a20504f/item/package.nix#L5-8
I should probably open an issue about that... I wodner why they changed Cabal behavior between ghcjs and ghc | 14:35:47 |
| fgaz | In reply to @sternenseemann:systemli.org
I wodner why they changed Cabal behavior between ghcjs and ghc My guess is that the jsexe installation is gated behind the --ghcjs flag, which is only used for ghcjs | 15:01:16 |
| fgaz | If so it should be an easy fix | 15:01:27 |
 pwmosquito | Redacted or Malformed Event | 16:11:01 |
| pwmosquito | * ~Heretic question but anyone knows if there's a similar channel for rust? Got a simple rust project i'm creating a shell.nix for, all is well except that i'm targeting wasm via cargo build --release --target wasm32-wasip2 and not sure how to bring that in? (i'm a complete rust noob and with rustup i'd do rustup target add wasm32-wasip2 according to the readme)~ found it | 16:35:22 |
| alexfmpe | In reply to @sternenseemann:systemli.org
contributions can be based against a merge-base of master and haskell-updates for testing I have a couple fixes that get non-TH libs building on pkgsCross.aarch-android-prebuild.haskell.packages.ghc910 based on current haskell-updates, do I open PR or just wait until things un-draft?
fwiw with ghc98 something is borked in the triple parsing and fails, but ghc folks confirm configure looks right so it looks like a bug in the rts configure stuff or something | 19:13:04 |
| alexfmpe | In reply to @sternenseemann:systemli.org
contributions can be based against a merge-base of master and haskell-updates for testing * I have a couple fixes that get non-TH libs building on pkgsCross.aarch-android-prebuild.haskell.packages.ghc910 based on current haskell-updates, do I open PR or just wait until things un-draft?
fwiw with ghc98 something is borked in the triple parsing and fails, but ghc folks confirm configure looks right so it looks like a bug in GHC's rts configure stuff or something that isn't present or triggered with 9.10 | 19:13:25 |
| alexfmpe | * I have a couple fixes that get non-TH libs building on pkgsCross.aarch64-android-prebuilt.haskell.packages.ghc910 based on current haskell-updates, do I open PR or just wait until things un-draft?
fwiw with ghc98 something is borked in the triple parsing and fails, but ghc folks confirm configure looks right so it looks like a bug in GHC's rts configure stuff or something that isn't present or triggered with 9.10 | 19:13:44 |
