fsprojects
DRAFT: paket move to FAKE-only
@webwarrior-ws can you try removing NU1605 now please?
@knocte where is it?
@knocte where is it?
https://github.com/fsprojects/FSharpLint/blob/58dd6b6704f2787c8c1dfdfa838ba28e29847dc5/Directory.Build.props#L9
You were lucky that I was not afk now :) but please, next time, for obvious questions use grep ;)
Build fails with multiple errors if I remove NU1605 from
<NoWarn>
Aren't those easier to resolve now that there's no paket?
If the last commit worked, then it shouldn't be marked as WIP anymore.
BTW when writing commit messages, if you are explaining 2 things that the commit does, please list them with bullets. Otherwise just listing each in a sentence, in the same paragraph, simply separated with a dot (with no way to join the sentences such as "hence", or something) feels kinda strange.
Good work wrt "build: re-enable NU1605" commit. Now, looking at it, it looks fairly simple. Are you sure we cannot separate into its own PR (aka: do it in master while it still paket-based)?
Good work wrt "build: re-enable NU1605" commit. Now, looking at it, it looks fairly simple. Are you sure we cannot separate into its own PR (aka: do it in master while it still paket-based)?
I tried, and it didn't work. Microsoft.Build.* versions are set by Visual Studio and cause package downgrades.
I tried, and it didn't work. Microsoft.Build.* versions are set by Visual Studio and cause package downgrades.
I never mentioned VS here. The topic is paket vs CPM (at CI level).
I tried, and it didn't work. Microsoft.Build.* versions are set by Visual Studio and cause package downgrades.
I never mentioned VS here. The topic is paket vs CPM (at CI level).
NU1605 was disabled because build in Visual Studio failed. See 58dd6b6
Note about <NoWarn>$(NoWarn);...</NoWarn>, I think it is almost always better to use <WarningsNotAsErrors>$(WarningsNotAsErrors);...</WarningsNotAsErrors> instead.
NU1605 was disabled because build in Visual Studio failed.
If you remove that NoWarn in master, then CI doesn't build. So we're talking about CI here, not VS.
CI failure seems to be a random bug - CI run for the same commit in cpm-rebased-2 branch is successful: https://github.com/webwarrior-ws/FSharpLint/actions/runs/7828237116