Skip to content

Conversation

@ErichDonGubler
Copy link
Member

Technically regressed from #8796; we forgot to cover all target directories we know we create.

@ErichDonGubler
Copy link
Member Author

CC @iwanders and @inner-daemons for the knowledge spread. 😄

@ErichDonGubler ErichDonGubler added type: bug Something isn't working area: infrastructure Testing, building, coordinating issues labels Dec 31, 2025
@inner-daemons
Copy link
Collaborator

This is on me, I told them to add the leading slash because I assumed we didn't create any other target directories.

Given that I doubt people will be symlinking the target directories in xtask, I wonder if we should just use:

/target
target/

That way, we cover symlinking the root target directory and don't regress other target directories, while still allowing files named "target"?

Also, does this cover naga's xtask?

@inner-daemons
Copy link
Collaborator

Also, @ErichDonGubler, it's the holidays! You've earned a break from working on wgpu lol, though your time is very much appreciated

@iwanders
Copy link
Contributor

iwanders commented Jan 1, 2026

Hopefully this is the only crate that's independent of the workspace? 🤞

That way, we cover symlinking the root target directory and don't regress other target directories, while still allowing files named "target"?

That would probably address the majority of usecases? When I manually write the .gitignore I usually just type target, which ensures that the build artifacts in crates not part of the workspace are also ignored. That does not match target.rs, but it would of course match a module directory called target, which would not be an unexpected name for the target of a rendering pipeline or something... I'm sorry this became such a rabbit hole 😅

Thanks for addressing this @ErichDonGubler, the branch name made me chuckle btw :D

@ErichDonGubler
Copy link
Member Author

I think it's pretty reasonable to expect any other main-workspace-independent crates to add their own .gitignore with /target, so I don't think we need a target/ rule specifically. OTOH, if anybody feels strongly about it, I don't expect it'll cause significant problems to add target/.

@ErichDonGubler
Copy link
Member Author

@iwanders:

Thanks for addressing this @ErichDonGubler, the branch name made me chuckle btw :D

Fun fact: You can find the source for this here:

https://github.com/erichdongubler-dotfiles/nushell/blob/61e46c9de3ba160280a115fc51cb3cb6d79ac024/scripts/erichdongubler/jj.nu#L177-L184

Copy link
Member

@cwfitzgerald cwfitzgerald left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I don't understand why this is happening. The xtask is part of the workspace, so will re-use its target directory.

@inner-daemons
Copy link
Collaborator

🤦‍♂️ I didn't even check that lmao. I assumed at least naga xtask would use a separate cargo due to being run from a different root folder but it seems cargo is still smart enough to use the root target. Unless someone is actually experiencing this issue probably this can be closed

@ErichDonGubler
Copy link
Member Author

Ahhh, derp, the target/ folder I was reacting to was left over from my previous development of xtask before it got added to the workspace. Closing!

@ErichDonGubler ErichDonGubler deleted the erichdongubler-push-swaggering-considerate-peach branch January 3, 2026 19:36
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

area: infrastructure Testing, building, coordinating issues type: bug Something isn't working

Projects

None yet

Development

Successfully merging this pull request may close these issues.

4 participants