>> I guess spawnl semantic (i.e, like win32's CreateProcess()) can't
>> become the default multiprocessing behaviour...
Nick> It would also make it much easier to write cross-platform
Nick> multiprocessing code (by always using the non-forking semantics
Nick> even on fork-capable systems)
I don't understand. On Unix-y systems isn't spawn* layered on top of
fork/exec?
One thing that nobody seems to have pointed out is that the subprocess
module was originally written as a multi-processing module with an API very
similar to the threading module. That is, it was intended to be used as an
alternative to threading. I would find it odd to use both together, and in
particular, to create threads first, then fork. If you were going to
combine both threading and multiprocessing it seems much more logical to me
to fork first (coarse-grained subdivision) then create threads (finer
grained threads of control) in those processes.
Skip
_______________________________________________
Python-Dev mailing list
Python-Dev@python.org
http://mail.python.org/mailman/listinfo/python-dev
Unsubscribe:
http://mail.python.org/mailman/options/python-dev/archive%40mail-archive.com
