On Thu, 4 Feb 2010, Randal L. Schwartz wrote: > Which was true in the 70s, and has no applicability today. Please stop > perpetuating the myth.
Er, ... may we disagree about this issue? Many of the python functions in mathematical libraries such as NumPy and SciPy are re-written in Cython. Cython produces C code that is compiled to shared libraries for speed of execution. I suspect that we're both correct, Randal, depending on the application. Those that tend to be I/O bound probably run as quickly when interpreted as when compiled, but serious number crunching tends to be CPU-bound and compilation makes a big difference. Rich _______________________________________________ PLUG mailing list [email protected] http://lists.pdxlinux.org/mailman/listinfo/plug
