> 
> >very interesting.  do you have any numbers for dynamic versions?
> 
> i did include dynamic numbers ... but i don't guess i labeled it very well,
> sorry.  I used the command:

oops; i didn't read it carefully enough!

> let me re-state the results (the number is dhrystones per sec):
> 
> 		dynamic		-static
> CASE1		4940.45		5010.19		! single usr, cache disabled
> CASE2		36727		37247.4		! single usr, cache on
> CASE3		24138.7		37170		! multiuser, cache on

that's an interesting result.  i'd not have expected dynamically
linked binaries to perform better, in general, and given that they
did, i'm surprised that in the multi-user test the statically linked
binaries performed better...

"weird."


chris