New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
deno 1.7.0 #66920
deno 1.7.0 #66920
Conversation
Is it possible for this not to be merged until after #66285? The CI run for that PR takes nearly two full days to complete, and this PR will create a merge conflict. If I have to rebase the Rust PR against master again, then that will cause another two-day delay. |
Removed the patch in PR, which is included in deno Got no problems to wait. Up to mods to respect your request @carlocab 👍 |
It would help if you marked this PR as a draft in the meantime, but I hope a maintainer sees my previous comment before approving. |
Will do. Thanks! |
Can |
From that issue:
This formula does build V8, if I'm reading it correctly. I suggest wrapping up this PR first, then you can open a separate one to do the migration. |
The formula itself doesn't, but |
Rust PR merged. Thanks for waiting, @chimurai. |
Thanks all, merging! :) |
|
Seems to succeed on 10.14 but not on 10.15 and 11.0. |
Sorry, I think this will have a conflict now that #67043 is merged. I'll resolve and update here. |
This is weird:
|
Shall we try to build if for Apple Silicon as well next time? |
Yes. I'd love ideas on how to fix the Catalina build tho. |
The error is in the gn / rusty_v8 part and at about 60% through the log: for 10.14
for 10.15 (old run):
for 10.15 (new run):
You have to scroll up through a lot of rust output to get to the actual gn output with the errors. |
I think those come from my having removed the Xcode build dependency. I couldn't find any documentation about it being needed, whether in Deno's or rusty_v8's docs, so I thought it would be safe to remove. |
I'm pretty sure the V8 build has some hard coded requirement on XCode in it's build system and doesn't support CLT only systems. |
I believe that; it's just not mentioned in their docs. (They even explicitly only require CLT, and mention that you don't need it if you have Xcode.)
Superenv changes environment variables depending on the declared dependencies though. For example, without the Xcode build dependency,
With it, it looks something like
These environment variables then affect how the |
But this doc are for people who use their prebuilt (rusty_)V8 binary and not accurate if you build all dependencies, including (rusty_)V8 from source.
Okay, that might be the cause then. It's still strange that the build failures aren't deterministic and seem to fail randomly in different steps. But maybe the CLT |
Ok, looks like 10.15 failed in 77e4ad8 long before it decided to stop: https://github.com/Homebrew/homebrew-core/runs/1741214532#step:7:6447 I wonder if there's some sort of |
Yup, I had a look at the rusty_V8 docs too. No mention of Xcode. Though, I suppose this isn't as carefully documented if they just fully intend for you to download a binary. |
Yes. See #66920 (comment) |
Yeah, the complicated why Chromiums tries to find the macOS SKD is here: |
Talk about non-deterministic. Compare: 54c027e with CI logs at https://github.com/Homebrew/homebrew-core/runs/1745216015 vs 77e4ad8 with CI logs at https://github.com/Homebrew/homebrew-core/runs/1741214515 Is it just me, or are these two identical formulae with exactly opposing outcomes? |
Ok, I patterned this latest commit after carlocab/homebrew-personal@0ab0b25 which builds successfully on GitHub runners (see https://github.com/carlocab/homebrew-personal/runs/1744816923).
|
Ok, let's try to be at least a little systematic about this.
|
Ok, the current version of the formula builds on:
Just not all at the same time. I guess if we just keep re-running CI, we'll get lucky and find a run where all of them succeed? Otherwise, I may try and merge a run where ARM and Mojave build, and just try to bottle for Catalina and Big Sur separately. (Not sure if |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I'm going to try to merge this and bottle for Big Sur separately.
Big Sur bottle job: https://github.com/Homebrew/homebrew-core/runs/1753554159?check_suite_focus=true 🤞 |
Second bottle atttempt succeeded: 5c23efe Thanks for the help all! |
Non-deterministic builds reported at denoland/deno#9244. |
Created with
brew bump-formula-pr
.