Skip to content

Conversation

@Caleb-T-Owens
Copy link
Contributor

No description provided.

@vercel
Copy link

vercel bot commented Mar 31, 2025

The latest updates on your projects. Learn more about Vercel for Git ↗︎

Name Status Preview Comments Updated (UTC)
gitbutler-components ✅ Ready (Inspect) Visit Preview 💬 Add feedback Apr 1, 2025 4:31am
gitbutler-web ✅ Ready (Inspect) Visit Preview 💬 Add feedback Apr 1, 2025 4:31am

@github-actions github-actions bot added the rust Pull requests that update Rust code label Mar 31, 2025
@Caleb-T-Owens
Copy link
Contributor Author

FYI @Byron

There was an issue where the parents of the workspace commit are not always up to date.

While this is something we seperatly need to fix, it also makes sense to move to something more reliable. (And also reduce the amount of forking in the function). As such, I've switched it to always use the multi-branch case
@Byron
Copy link
Collaborator

Byron commented Apr 1, 2025

The test-failure seems to be flake, and I have seen that too in other PR.
And since the removed lines don't trigger a legit test failure, I presume there is no reason to keep it. Further, I think with #7596 merged, the handling of 'framings' would be centralised and well tested.

Now I think it's time to fix the likely source of the issue, and finish GitoxideLabs/gitoxide#1851 which will remove parallel iteration support.

Copy link
Collaborator

@Byron Byron left a comment

Choose a reason for hiding this comment

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

While GitoxideLabs/gitoxide#1851 is waiting for CI to be merged, I am rerunning CI here and merging it assuming it will go through this time.
The likely gitoxide fix will then be coming in with #7902.

@Byron Byron enabled auto-merge April 1, 2025 03:23
This brings in a fix to the way references are sorted, assuring
they always come out in the right order.

Maybe this fixes the flakiness issue.
If not, maybe it's something else though?
@Byron Byron merged commit 5fbc50d into master Apr 1, 2025
18 of 19 checks passed
@Byron Byron deleted the fix-commiting-with-one-stack branch April 1, 2025 04:37
@Caleb-T-Owens
Copy link
Contributor Author

Hey @Byron,

Here is a bit of a fuller explanation as to why I made this change. (I wrote this explanation two times already and v3 deleted them so I gave up and just pushed).

We were getting what seemed to be this out-of-sync error. As it turns out, there are some cases where we're not calling update_workspace_commit which results on the parent's of the workspace commit not always being a direct analogue for the heads of the stacks.

While we obviously need to tighten up where we call update_workspace_commit, it also seems reasonable to always try and get the stack head from a more reliable location, which is the stack head.

It also removed a code branch which makes everyone happy.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

rust Pull requests that update Rust code

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants