search

ionide / FsAst

http://blog.ctaggart.com/2014/09/generating-f-code-using-its-ast.html
MIT License
48 stars 15 forks source link

Update to FCS 40 #36

Open baronfel opened 3 years ago

baronfel commented 3 years ago

We're blocked from doing this until fantomas has a version that supports FCS 40. There's a branch with that work, but no prerelease/stable quite yet.

Zaid-Ajaj commented 3 years ago

Wouldn't be a good idea to keep FCS used here in sync with FCS version in Fantomas? Had problems using different versions

Heck, I even thought about including FsAst into Fantomas (maybe in a .Writers namespace) since they are always used together

baronfel commented 3 years ago

Wouldn't be a good idea to keep FCS used here in sync with FCS version in Fantomas? Had problems using different versions

yes, that's precisely why the associated MR is in draft state still :p

baronfel commented 3 years ago

@nojaf speaking of, any idea when the first prereleases of fantomas with FCS40 will drop? I'm also looking forward to those for FSAC itself

nojaf commented 3 years ago

Hey, yeah about that 🙈. I'm currently pitching the idea of having a patch release for FCS 40. There is one thing that just really doesn't produce AST anymore. And another fix that would allow for a huge clean up in Fantomas. I've poked @vzarytovskii about it and they would discuss it internally.

baronfel commented 3 years ago

the thing is, even that proposed patch release would be a major version bump (due to api breakages). so it sounds like you're suggesting skipping 40 entirely?

nojaf commented 3 years ago

It has crossed my mind, to be honest. The daemon experiment could potentially resolve the hurdle in FSAC. And as for FsAst, I'm not sure what the impact would be. If there is any...

nojaf commented 3 years ago

https://www.nuget.org/packages/FSharp.Compiler.Service/40.0.1-preview.21352.5 just dropped, I'll check over the weekend if I can publish the first alpha.