On Sun, Jul 29, 2012 at 5:52 PM, Andrew Berg <bahamutzero8...@gmail.com>wrote:
> On 7/29/2012 7:12 PM, Rodrick Brown wrote: > > Python is a glue language much like Perl was 10 years ago. Until the > > GIL is fixed I doubt anyone will seriously look at Python as an option > > for large enterprise standalone application development. > The GIL is neither a bug to be fixed nor an inherent part of the > language. It is a design choice for CPython. There are reasons the > CPython devs have no intention of removing the GIL (at least in the near > future). A recent outline of these reasons (written by one of the > CPython devs) is here: > > > http://python-notes.boredomandlaziness.org/en/latest/python3/questions_and_answers.html#but-but-surely-fixing-the-gil-is-more-important-than-fixing-unicode It's a nice document, though it seems to use the phrase "shared memory" in a novel (to me) way, and literally says that multiprocessing doesn't use "shared memory" even though it does (at least in the sense of the phrase that I'm accustomed to). I suppose you could call what I usually refer to as "shared memory", instead "System V shared memory". It's hidden from the user to a large extent, but when multiprocessing passes objects from one process to another, I believe it's doing so via System V shared memory.
-- http://mail.python.org/mailman/listinfo/python-list