@sandydoo since this is touching Haskell modules (which is indeed the right way to do it!) I think we might need to first merge it to haskell-updates branch.

@sternenseemann or would it be OK with you if we merge this patch directly to main?

Oh, that did not do what I wanted 🤦 Let me fix this...

gosh. Yeah now the diff is all the master commits that haskell-updates lack behind. Version control is fun. Maybe you can keep master as base for now. I think maybe @sternenseemann will be OK with us merging this to the master? I'm not sure about how many rebuilds that could cause.

If not then maybe new branch based on haskell-updates and cherry-pick the comment on top of that? That's probably what I would do.

Waiting for the the latest branch to download 🙂 Sorry for the noise, everyone!

Oh no worries at all. I'm so glad someone picked this up. I was afraid I will need to steal Mac from my GF and set nix and everything up myself. Thanks a lot for looking into this.

Also... If you wish -- consider adding yourself to the maintainers of elm. It would be great to have someone with MacOS at hand there. It's important platform we need to keep supporting.
Just reminder that this is usually done in its own commit so if you choose to it can be its own PR.

Managed to wrangle git. Should be good now.

Also... If you wish -- consider adding yourself to the maintainers of elm. It would be great to have someone with MacOS at hand there. It's important platform we need to keep supporting.

Will do, thanks!

This can't go to master due to the high amount of rebuilds. I'm also confused – weren't we supposed to be able to build security tool again?

I'm also confused – weren't we supposed to be able to build security tool again?

So this just applies the same workaround that was used for x509-system. A lot of packages, including http-client-tls have switched over to crypton-x509-system, which inherits the same issue with security.

I can try out the security-tool approach tomorrowsome time. Has anyone managed to build it?

Can we merge this?

I think so. I think usually maintainer of haskell-updates does so but this is already open for some time.

Successfully created backport PR for staging-24.05:

• [Backport staging-24.05] elmPackages.elm: fix build failure on darwin #318352

I'd still appreciate it if someone could look into packaging the tool properly so we can get rid of these unsatisfactory workarounds!