Hi Osvaldo and List,
I had to translate it by hand. It's not hard - I don't know OPS5
but it was easy enough to figure out the translation.
Since the files are small, I'm going to break mailing-list protocol and
include them here as two attachments. You have to compile the C program
and use it to generate test data for the Jess program. Then you have to
edit manners.clp to specify the name of the data file (near the end).
One more note: no currently released version of Jess will properly run
this benchmark! (only my working version will.) The reason is that Jess
has
historically been lax about relative ordering of near-simultaneous rule
activations, but this benchmark requires precision in this area. The
next
Jess release will include this new precision.
I think Osvaldo Pinali Doederlein wrote:
>
> > CLIPS Jess
> > ------------------------------
> > manners/32: 5 5
> > manners/64 68 22
> > manners/80 235 56
> > manners/90 456 90
> > Note that this is fully twice as fast for manners/90 as my last
> > reported time. Most of the gain is due to HotSpot; about 15% is due to
> > further tuning of Jess.
>
> Hmmm where can I find this benchmark code? :) I searched but only found for
> OPS5...
>
> Thanks
> Osvaldo
---------------------------------------------------------
Ernest Friedman-Hill
Distributed Systems Research Phone: (925) 294-2154
Sandia National Labs FAX: (925) 294-2234
Org. 8920, MS 9214 [EMAIL PROTECTED]
PO Box 969 http://herzberg.ca.sandia.gov
Livermore, CA 94550
manner_data.c
manners.clp
