Hey! I want to run an app in a Rust runtime. This app requires access to a Rust toolchain and a CLI tool which must be installed via a curl command. Currently, I’m unable to access the CLI tool when running the “start command”, causing my app to fail.
The Rust caching in Render is fairly straightforward and doesn’t require manual modification which Docker requires in order to improve caching. Why is it not recommended to use the Rust runtime?
I think I misunderstood the issue, I still don’t understand why sp1 isn’t in Crates (though I did find zkp there, and sp1’s site says “Use SP1 or any proof system to write a ZKP for your application.”), but I have no active knowledge in Rust to inform why one would use zkp, sp1, etc.
In short, I believe the issue here is that the execution of the Build Command, Start Command, and even processing of Environment Variables isn’t POSIX, or more specifically, it doesn’t expand variables. I believe what you want to do is to throw those lines in a build script, probably headed with #!/usr/bin/env bash, which should then allow you to stuff things like $PATH changes into processing, then set your Build Command as ./build.sh or something of the sort.
I believe that adding the path change into ~/.bashrc will then persist in the environment’s boot process, leaving you to only need to run cargo prove --version as your Start Command.