Carl Friedrich Bolz wrote:
[Snip]
> First let me note that I think the extcompiler is a promising idea. I
> agree that the idea could lead to something generally useful. However:
> The current implementation is not useful at all. It wasn't used
> seriously by anybody. Richard tried to use it and he found that it
> wasn't creating anything useful, since the resulting CPython
> extensions weren't really any faster than just writing the stuff in
> pure Python to begin with.
> 
[Snip]

I don't have the experience, time or motivation to maintain the 
extcompiler so I completely understand if you need to remove it. But I 
do want to say that there are situations where it is useful and a speed up.

My use of the extcompiler was making a mixed module that is a board 
representation and move generator for a board game (Arimaa). The 
resulting extcompiler module was approximately 10 times faster than the 
corresponding regular python module. So I think the ability to see speed 
increases with the current extcompiler is probably heavily dependent on 
how frequently the cpython to rpython interface has to be crossed.

Janzert

_______________________________________________
[email protected]
http://codespeak.net/mailman/listinfo/pypy-dev

Reply via email to