[IronPython] Re: Nice speed

Michael Spencer mahs at telcopartners.com
Wed Apr 27 04:15:08 CEST 2005

Travis Watkins wrote:
> Never trust the program you're running to give you accurate time. 

I thought (and think) that the measurements are sufficiently accurate, and the 
differences sufficiently marked to support the speed claim.  I also ran the test 
with much longer cycles, checked the reported time with wall-clock time, and 
obtained comparable results.  Repeating each test several times, and discarding 
outliers reduces the impact of other processes.  What sort of inaccuracies are 
you concerned about here?

Of course, this is far too narrow a test from which to draw any broad 
conclusions about the speed of the implementation.

> 'time python foo.py' instead.

I presume you mean on ~nix.  I'm using Windows, and I'm not aware of a similar 
facility (which doesn't mean it doesn't exist).

  If you're going to do that put most of
> the code in a seperate file that gets imported so the bytecode gets
> cached since IronPython is precompiled to IL. 

I compared pre-compiled functions in each case.  Can you explain why what you 
suggest is fairer?

Other than that, cool.
It is, isn't it ;-)


More information about the Ironpython-users mailing list