Unable to run without spawning at least 1 native thread

[4xrsJCr9]

Unable to run without spawning at least 1 native thread

[dpc]

Right. The first mioco thread should be the current one.

[dpc]

Is this really a big issue for you? I looked at the code, and the biggest reason to do this the way it is currently is simplicity.

[4xrsJCr9]

It isn't a big issue. My primary concern was related to the less than ideal interaction between threads and signal handlers, but that can be worked around.

Thanks for looking into it.

[Drakulix]

Its actually an issue for me. Many ffi-libraries rely on having access to the main thread. With mioco this is impossible, as the main-thread will be always joining all worker threads, instead of becoming a worker thread itself.
I think it would be quite easily to implement by moving the thread creation out of spawn_thread into start (and probably renaming spawn_thread in the process to something like thread_loop). Then start could call that function itself at the end.
Would you interested in a pull request for this?

[dpc]

Sure.