Can you share more about how you’re trying/hoping to use it? Do you have something that already uses some kind of cloning and you’re trying to improve how you use it?
I’m building a distributed background job processing with machines as workers. I leverage the Machines API for this purpose. I’ve found many things that work well when using flyctl that do not work the same when using the Machines API. Ideally, both should result in the same result.
In this case, it’s not even possible to clone machines through the API, there’s no endpoint for it. Creating machines via the API is cumbersome enough as you have to send the whole configuration in the request body (why can’t it just read it from the app’s configuration as when you do fly scale count?). So I keep around a set of stopped worker machines on standby, but if there’s a demand spike, I need to create more on the fly. Cloning from the existing machines would make this process simpler.