your jsi_codegen.yml is different from master, not sure why gate is failing.

Apply feedback

otherwise looks good to me

Strip out all unneeded defaulting logic

Add comment about OSX setting

I think he means manually

This is really helpful context, would've been great to mention this at the start.

During install:

$ curl --proto '=https' --tlsv1.2 -sSf https://sh.rustup.rs | sh
...
This path will then be added to your PATH environment variable by
modifying the profile files located at:

otherwise LGTM

This works, but the output is different than original:

Remove leading './'

Fair enough, but it was nice to be able to iterate on these in context of a PR.

Since people have to opt-in to this and they have to have direnv pre-installed, I think this is fine for now.

I was just following what rustup does, as I assume they have had many years of refining what files need to be written to.

This seemed to have worked the first time without issue. I would to avoid the order of install targets determining whether a pipeline is green.

We're moving these out from native/android/app/CMakeLists.txt, yeah?

What are the pros/cons of using our existing install_protobuf.sh script vs. arduino/setup-protoc@v1? Kind of weird to use an Arduino-specific GH Action, but I presume install_protobuf.sh isn't working here because of the make install?
Ultimately defer to you and @atul, who have spent the most time thinking about this.

Address feedback

Rubber stamping as there's no "smells", would like to eventually see a commtest around this

Rubber stamping as there's no "smells", would like to eventually see a commtest around this

We should probably rerun the build if the protos or lib.rs change

LGTM, I'll try this on the "ensure" nix work I'm still churning through

rebase should fix CI gate

There's likely a more idiomatic rust way to do this. But it escapes me (and c++ interop makes some conventions awkward).

going to rubber stamp this,

Gate should be fixed if you rebase on master