On 31/01/2011, at 2:15 PM, Jeff Schultz wrote:

> On Mon, Jan 31, 2011 at 07:29:41AM +1100, john skaller wrote:
>> Not so good:
> 
> Indeed.

Of course  it is probably a tad faster than Python :)

Be interesting to see how a "native" C++ program fared.

> Coding the same thing with a Clojure-style sequence library
> JDK1.6.0_23 gets
> 
> real    0m8.112s
> user    0m0.015s
> sys     0m0.000s
> 
> (Cygwin /bin/time only really reports elapsed time in this case.)
> 
> If ocamlopt is only twice as fast as a quite inefficient Java version,
> there must be something it's missing!
> 
> 
>> Reality is worse: parameters set in Felix so there's no call to the GC.
> 
> Amusingly, giving the JVM *more* memory to work with than its default
> slows it down significantly.


I watched the Felix gc. When there's a few huge lists, it is very VERY slow ..
like "hardware reset" slow :)

When the memory size goes past my Mac's 2G Ram it gets into paging spasms :)

The good news is that even for huge numbers and pointless collections,
the Felix program didn't crash :)


--
john skaller
skal...@users.sourceforge.net





------------------------------------------------------------------------------
Special Offer-- Download ArcSight Logger for FREE (a $49 USD value)!
Finally, a world-class log management solution at an even better price-free!
Download using promo code Free_Logger_4_Dev2Dev. Offer expires 
February 28th, so secure your free ArcSight Logger TODAY! 
http://p.sf.net/sfu/arcsight-sfd2d
_______________________________________________
Felix-language mailing list
Felix-language@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/felix-language

Reply via email to