-
Notifications
You must be signed in to change notification settings - Fork 8.9k
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
Rewrite git history #31562
Rewrite git history #31562
Conversation
Good eye @jnewb1 - yes, it's very small ❤️ GitLab LFS is 6.3 GiB though, but that's not an issue. Since when you cloned openpilot-tiny, only 250MB of LFS files are downloaded. I uploaded the LFS on GitLab (as you do). One reason I keep the last lines in the script commented, is that it spams LFS and will upload LFS files everytime the reference ids of the lfs files change. So, don't run the script until I finish it. |
scripts/bfg.sh
Outdated
# /tmp/git-filter-repo --invert-paths --path external/ --path phonelibs/ | ||
|
||
# check the git-filter-repo analysis again - can be found in the repo root/filter-repo/analysis | ||
/tmp/git-filter-repo --force --analyze |
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 most interested in knowing the top remaining files and their sizes. does this have it?
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.
now you can find the latest analysis in this folder https://github.com/andiradulescu/openpilot/tree/bfg/scripts/analysis - I will try to import more files to lfs and I will keep updating the folder.
fyi external
and phonelibs
folder were initially deleted completely at the time of this comment #31562 (comment) - so still need to either add both folders to lfs or just add the big files inside them
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.
analysis of the current commaai/openpilot repo can be found here https://github.com/andiradulescu/openpilot/tree/bfg/scripts/analysis-before
My strategy (like it or hate it):
The strategy takes 30 days to apply, but shouldn’t disrupt anyone in the 30 days. After the 30 days, will be disruptive for users that need to move forcefully on main. Didn’t do this strategy before. It’s a new idea. |
If I can add some input, for the step 6 I’d only apply it to new PRs, otherwise it gives a 30day timeout for existing WIP which might be a bit tight. New PRs after the deadline should be rejected in favor of PRs to main |
RE: rollout, I'm not interested in dealing with a slow rollout, so we'll do it quickly and put some effort into rebasing PRs automatically for people. |
Small update here: rebasing Right now, I'm testing this since rebasing 12040 commits is not that fast. And during the rebase I get some logs like this: |
Anything I can help with on this? Is there a particular blocker? |
The current blocker is Otherwise, you can unlock the bounty. Maybe @devtekve is interested. |
Thanks ❤️ but not really, if it’s for me don’t unlock it! I was just chiming in. If there’s anything I can help with feel free to ping me tho |
Thanks for contributing to openpilot! In order for us to review your PR as quickly as possible, check the following:
|
@adeebshihadeh I edited the first comment with some checkboxes - #31562 (comment) - let me know what you think |
Merging as is, feel free to open a pr if you need to add something to your script. We still have to wait untill all the cars stuff is out before actually doing the rewrite |
Glad seeing this merged! I’m still running it and changing small things to it, I’ll open a new PR after I stop testing and polishing. |
This tackles #28840 and is a continuation of #30824.
This PR addresses
master
:v0.7.1
git cherry-pick A^..B
where A is the first commit and B is the last commit, from their old PR)git push
failsContinued in #32955