>import random
>import time
>
>a =0
>t0= time.clock()
>for x in xrange(1, 10000000):
> a = a + random.random()
>print time.clock()
>
>Linux - 2.96 (avg for 10 runs)
>windows xp - 2.28 (avg for 10 runs)
>Mac - 3.33 (avg for 10 runs)
>
>I have not a clue as to why the Mac is so slow. But in general these types of timing test of language vs language do not prove much even on the same box. That said I have not found much difference between VFP and Python in general speed. I do believe is there is a difference between python GUI programs and VFP. VFP loads faster. But the actual execution of the code I have never noticed anything.
>
>It's also true that the speed of loops is faster on the newer versions of python (3.x). But I doubt it matters much.
>
>What I think is important was the python code is readable (even looks like your VFP code) and most important runs on 3 platforms.
>
>Johnf

True, but my only point is that there are things to be said for VFP being a MS product for so long. One of them is that it's more of a first class citizen in the Windows world and will have a lot of intangibles going for it. My point is: if I'm doing Windows (and not other platforms) I would choose VFP or C# over Python, Ruby, PHP because VFP/C# is always going to have an edge on Windows. I'm willing to bet that if VFP loses any performance test to those languages on Windows then there is some sort of work-around that will make VFP the winner.