🚀 go-pugleaf

In article <1993Jan18.142453.1@vaxa.strath.ac.uk>, cnbv01@vaxa.strath.ac.uk writes:
> Excuse my ignorance, but is there any efficiency advantage to having modules.
> 
> I have developed, what I would call, a large application in Quintus
> Prolog which is spread over many files.  I would like it to run faster...
> now thats a new one!!...and am presently checking for determinacy in the
> rules.  (I'm just realising all the beauties of Quintus Prolog 3.1.1 on
> a DECstation after years of version 2.4 on a VAX.)  In addition I've
> heard and read alot about modules... and I have not used modules ie. all
> my rules are in module user.  
> Is there any benfit to having modules?...Are rules called faster if they
> belong to modules with fewer rules in them?
> 
> ...any ideas
> 
> Thanks,
> Graeme Burt

Sorry, there is no performance advantage in using modules.  The purpose
of modules is to divide up the predicate name space, so that you don't need
to worry about inadvertently using the same name for two different predicates.

Getting any unintended non-determinacy out of your program could be a big win.
library(det) is a crude but useful static analysis tool to help with this.
Using the debugger and skipping through sections of your program can also be
a good way to find non-determinacy.

In a future release there will be a profiler, which I know many people will
find useful for identifying performance bottlenecks.

-- 
David Bowen				Vice President Product Engineering
Quintus Corporation                     InterNet:       dbowen@quintus.com
2100 Geng Road                          FAX:            415 494 7608
Palo Alto, California 94303             Phone:          415 813 3800