[Python-Dev] Threading in the Standard Library Tour Part II

[Tim Peters]

I agree it would help to say more about threading pragmatics.  It's
hard to say anything true, though <wink>.  I'll just pick on this:

[Raymond Hettinger]
> ...
> And, its overhead decreases total application performance.  Also,
> multiple processors cannot boost performance because Python's
> Global Interpreter Lock (GIL) precludes more than one thread
> from running in the interpreter at the same time (this was done to
> simplify re-entrancy issues).

CPython's threading was really designed to allow I/O to proceed in
parallel with one thread of Python computation.  If used for that
purpose, it's not true that total app performance decreases, nor is it
true that performance can't improve on a multi-CPU box.  Performance
can increase significantly in that case, and that's the prime use case
for *practical* threading in Python (that's why the GIL is released
around virtually all C-level I/O calls -- I/O concurrency is the use
case Guido cared about).

Given experience with Zope and ZEO, I'd also add about 10 pages
explaining that trying to mix asyncore with threads is legally
sufficient evidence of insanity <0.9 wink>.