Hi all, I joined the discussion here since last week and I want to help push the progress on this awesome parallel-rustc feature. As I mentioned in some previous messages, I still need some more information to evaluate current status of parallel-rustc.
The core problem I want to figure out is: "Is the feature mature enough to be used without significant performance regression?"
If the answer is yes, then of course we can move on to the following experiment phases.
If unfortunately it is not, then I want to know what is the main obstacle. And is there a plan to fix it?