Stackless Platform Independence?

[paul rubin]

Christian Tismer <tismer at tismer.com> writes:
> Both approaches make sense to me. On systems with no threads
> at all, we have only one choice, Paul's, unless we code
> something thread-alike for the specific platform.
> When real threads are available, the other is more powerful.

I like the idea of using microthreads when possible and resorting to
OS threads for legacy extensions that do blocking i/o.  There should
be a way to code new extensions to use nonblocking i/o and let
Stackless know that it's ok to call those extensions from
microthreads.