On Sun, 08 Jan 2006 10:24:43 -0500, Peter Hansen <[EMAIL PROTECTED]> wrote:
>[EMAIL PROTECTED] wrote: >> yes, I read it, and I even know about threading's existence. I just >> thought that if something claims to be atomic, it better should be. > >I think the term "atomic" is meaningful only when the context is known. > For example, "atomic" operations in the Python interpreter are >certainly not atomic within the larger context of the CPU, and atomic >CPU operations are not necessarily atomic in the context of a system >with multiple CPUs. If the context for mutex.py explicitly excludes >multi-threading then you have to interpret "atomic" in whatever context >that defines. > "Atomic" means trademarked by a company that used to use that name in the 50's to describe and identify a line toys it put in its breakfast cereal boxes. The rights are now owned by an IP scavenging company which is trying to sell them for stock in another IP scavenger with more money left, so be careful. There's also some related talk of a patent on a method of using semiotic elements in association with the distribution of digital generic products to make them distinguishable for any business purpose whatever. IP principles established with corn flakes and decoder rings are thought to translate perfectly to the digital domain of FOSS distros including anything toy-like. Regards, Bengt Richter -- http://mail.python.org/mailman/listinfo/python-list
