This PR adds an extra closure passed to addCallback, but it makes the continuation (identName) a nimcall in return, except if the async proc has FutureVar parameters. We can make an object of futureVars and pass it around to improve that case, I think. Maybe it's possible/better to move the futureVar logic to the closure iterator. That would also help to get rid of both closures like it's done in chronos, but that also requires a breaking change in addCallback.

That would also help to get rid of both closures like it's done in chronos, but that also requires a breaking change in addCallback.

Fine with me as long as there is some migration period and how to patch client code can be explained easily.

Thanks for your hard work on this PR!
The lines below are statistics of the Nim compiler built from f29234b

Hint: mm: orc; opt: speed; options: -d:release
178145 lines; 8.641s; 653.828MiB peakmem