use ocaml-css-parser lib#93
Conversation
|
@tatchi Nice!!! I'll review this soon, I'm super excited to up the OCaml version. EDIT: I think I know which test is failing --- see the fix I implemented for #89 . Those changes are not upstream in |
Cool, let's wait for a new release of |
|
@dylanirlbeck I bumped |
|
I think the error comes from the comparison of the |
|
@dylanirlbeck it builds now on windows but CI fails at a later stage: |
|
@tatchi Hmm I'm a bit lost on this one, especially because the test passes on Linux/Mac. I'll continue to investigate.. sorry for the delay on this, just started a new job haha |
|
@tatchi So I'm still stuck on this, and have posted on Discord in hopes of getting a response. What causes even more confusion is the fact that the same test passes in this PR, with the only difference between the two PRs being yours changes the OCaml version. I'm very much confused.. as a sanity check you could drop the esy version on CI to 0.6.2 and see what happens? It might be worth a shot. |
|
@tatchi See esy/esy#1096 (comment). Can you try the fix proposed in the comments by Manas? |
|
@dylanirlbeck looks like it fixes the issue :) |
|
@tatchi 🔥 🔥 🔥 I'm going to double-check to see if a new esy version is on the horizon, since I'm not sure if using the nightly build is super reliable -- just hang tight, I'll merge this in either way by the end of the week. |
Nice :) Not sure what are you plan about #100 but once this one is merged, I can rebase #100 and there won't be any needs for nightly version of esy |
|
@dylanirlbeck esy 0.6.5 was just released so I switched back to the latest version (not the nightly one). I guess it should be ok :) |
dylanirlbeck
left a comment
There was a problem hiding this comment.
Overall, these changes LGTM! I’m going to go ahead and merge. The BS-specific stuff is above my head, but since it was taken from graphql-ppx I’m sure it’s fine!
Fixes #87
I was finally able to use a higher version of OCaml (4.9.1) and still have errors formatted properly as before (thanks to @jchavarri). The trick was to use the
--embed-errorsoption.There's still a native test which fails I'm not sure why...
Note: this PR is blocked by astrada/ocaml-css-parser#6