Re: CCL:Benchmarks and Software

[Peter Shenkin <shenkin-0at0-schrodinger.com>]

Hi,
 I quite agree with a number of points that have been raised,
 including Dave's.
 But I feel that people should be able to publish benchmarks
 performed with their software so that prospective users
 can then decide (1) whether those particular benchmarks are
 relevant to their prospective uses, and (2) whether speed
 matters at all.
 What has only been implicit in all this is that some of the
 providers of softare disallow, in the standard license agreement
 which every user must sign, any publication of benchmarks.
 This, I feel, is inimical to the further development of the entire
 profession -- in fact, both professions (plural): computational
 chemistry and software development in this area.
 	-P.
 On Thu, 22 Apr 1999, David Young wrote:
 >
 > Hi,
 >
 > 	I would like to point out that this conversation on benchmarks is
 > a bit one dimensional.  There are some very popular (profitable) chemistry
 > software packages that do poorly on benchmarks but sell because of other
 > factors such as ease of use or functionality.  ...