picked this back up and turned it into a tool that can estimate within 10% the compile cost of functions. Think of it like cargo llvm-lines that rolls up to functions in your crate
xav
prototyped a little tool here to break down the sizes of large futures.
My objective is to help figure out how to reduce compile times in async-heavy projects by telling you which futures are worth erasing. Might work on surfacing call-graph depth next.
xav
Yeah, most of these things are *avoidable* in practice: making sure you don't have large owned items in futures, try to keep nested future depth low, etc.
It's just annoying you HAVE to work around papercuts.
