July, 2026: Google copybara allows one to move code between two prod repositories
March, 1974: IBM COPY allows one to move code between two prod partitioned data sets: OS/MVT and 0S/VS2 TSO Data Utilities COPY, FORMAT, LIST, MERGE User's Guide and Reference https://www.computinghistory.org.uk/downloads/8987
If the only need you need is sync repos without exclusions or transformations I wouldn't bother, it could work for you until it doesn't when they archive it or kill it like kaniko or so many other google products/tools.
Gitlab has really simple way to mirror from Gitlab to Github or other git vendors/servers
Been using this for a while, mostly when I make a tool as part of a larger project and the tool is big enough to deserve its own release.
It’s powerful enough to do a whole bidirectional shipping operation where you export and import code—no thanks, that’s a hassle. I use it mostly for a simple fire and forget export, where I take a folder out of its original repo and preserve the history. Then I just move development to the new repo. The new project layout can be completely different, but Git blame works and I’m happy with that.
The one-way pattern is actually how Google uses it internally too, syncing outward from their monorepo to GitHub. Bidirectional gets messy because transforms (path remapping, file exclusions, header stripping) are easy to apply in one direction but can't always be cleanly inverted. When both sides have diverged, Copybara's baseline tracking starts producing confusing results because semantically equivalent commits generate different SHAs after transform.
One thing worth knowing: history "preservation" is actually cherry-picks with rewritten commits, not a true transplant. Git blame works because the file content and authorship carry over, but the SHAs are new. Copybara embeds the original SHA in a commit message trailer (GitOrigin-RevId), which is useful to know if you ever need to correlate commits across repos after the fact.
To those who have used it: is it handy for situations where you have multiple repos that want to share a little code, but it's not worth the trouble of extracting a library, referencing it, publishing versioned releases, updating dependent repos, etc?
And instead just "sync" a code folder from one main repo (perhaps containing common domain models) to other repos?
Basically the Go philosophy that a little bit of copying is better than a lot of dependency?
It’s largely used for syncing external open source projects with the monorepo. Policy is to require source code imports over built artifacts. Though you can get exceptions.
Some projects are also developed in the monorepo and exported via Copybara.
My team also uses it to version Starlark rule sets internally.
Source code imports versus artifacts really neither here nor there. Go is source code imports too.
The key part for Copybara is that Google will make changes to the OSS projects from within the internal repo and everyone else will make changes to the OSS projects.
It's for when you have a monorepo internally, and want to publish parts of it as open source to the world. They still need to live in the monorepo, so this is the solution.
Having a public repo as a dependency for your private corporate repo is a pain in the ass development-wise. Having a tree of such dependencies is a migraine.
Yeah, that's the fun part. Probably built first for exporting monolith slices to OSS, but the reverse direction is more interesting to me. Tracking an upstream or keeping a private fork in sync. That's what makes Copybara useful well beyond the monorepo use case.
Copybara can do that but I think it will be annoying and tedious to use it that way. More annoying than the problem of extracting a library or shoving some files in a separate repo.
Interesting. Anyone knows how this compares to using git submodules and subtrees?
I had used those to create separate repo for website artifacts while the same also remain plugged into the webapp dev repo. (Both sides remain modifiable and changes mergeable to the
other side.)
If you're exporting more than a few commits, I suggest using local repos (/path/to/.git) for both the source and destination. Otherwise it'll be quite slow.
i used this tool when i was at google, extremely helpful in open-sourcing things from google3 to github.
still, i'm glad to just directly develop on github now :)
July, 2026: Google copybara allows one to move code between two prod repositories
March, 1974: IBM COPY allows one to move code between two prod partitioned data sets: OS/MVT and 0S/VS2 TSO Data Utilities COPY, FORMAT, LIST, MERGE User's Guide and Reference https://www.computinghistory.org.uk/downloads/8987
Gitlab has really simple way to mirror from Gitlab to Github or other git vendors/servers
It’s powerful enough to do a whole bidirectional shipping operation where you export and import code—no thanks, that’s a hassle. I use it mostly for a simple fire and forget export, where I take a folder out of its original repo and preserve the history. Then I just move development to the new repo. The new project layout can be completely different, but Git blame works and I’m happy with that.
One thing worth knowing: history "preservation" is actually cherry-picks with rewritten commits, not a true transplant. Git blame works because the file content and authorship carry over, but the SHAs are new. Copybara embeds the original SHA in a commit message trailer (GitOrigin-RevId), which is useful to know if you ever need to correlate commits across repos after the fact.
And instead just "sync" a code folder from one main repo (perhaps containing common domain models) to other repos?
Basically the Go philosophy that a little bit of copying is better than a lot of dependency?
Some projects are also developed in the monorepo and exported via Copybara.
My team also uses it to version Starlark rule sets internally.
The key part for Copybara is that Google will make changes to the OSS projects from within the internal repo and everyone else will make changes to the OSS projects.
Having a public repo as a dependency for your private corporate repo is a pain in the ass development-wise. Having a tree of such dependencies is a migraine.
Say, to rebase upstream MySQL changes onto a fork in the monorepo (in a random, non-specific example)
https://josh-project.dev
The blog post from the Rust people:
https://blog.rust-lang.org/inside-rust/2026/06/04/how-josh-h...
Meta used to have an open source tool called fbshipit. But according to its open source repo they no longer use it:
https://github.com/facebookarchive/fbshipit
Any others in this space?
It has since been merged into git proper:
https://manpages.debian.org/testing/git-man/git-subtree.1.en...
https://docs.github.com/en/get-started/using-git/about-git-s...
My shell script definitely wasn't google scale tho!
For example altering commit author emails during sync
I had used those to create separate repo for website artifacts while the same also remain plugged into the webapp dev repo. (Both sides remain modifiable and changes mergeable to the other side.)
Thx.
It works great and I've seen many teams gain significant productivity when collaborating in a monorepo with public bits.
If you're even toying with an internal monorepo you owe it to yourself to give it a try.
I’m curious what downsides folks have experienced with this tool?
Any tips?