Steve Holden wrote: > The basic answer is that so far no developer has felt it worthwhile to > expend time on adding these optimizations.
Mainly because it's rare to find such constructs in anything except contrived examples ... Nearly every time you use a literal, it's being added to (subtracted from, etc) a non-literal. Adding such optimisations to Python may improve it's benchmark scores, but that's about it. The real gains are to be found in reducing function call overhead, attribute access, etc - and as a result there has been considerable effort expended in those areas. FWIW, there are some of us who do occasionally do bytecode optimisations, but we know it's just a bit of fun. I've yet to do any bytecode manipulation that's significantly improved performance. You're best of leaving such things to Pysco, and concentrating on algorithmic improvements. Tim Delaney -- http://mail.python.org/mailman/listinfo/python-list