-----Original Message----- From: Pavlos Christoforou [mailto:pavlos@gaaros.msrc.sunysb.edu] Sent: Monday, May 24, 1999 1:41 PM To: Jim Salmons Cc: Chris Petrilli; zope@zope.org Subject: RE: [Zope] ZOPE, Medusa and the ZD WebBench 3.0 test suite...

On Mon, 24 May 1999, Jim Salmons wrote:

...

Despite whatever we and Twain may think about benchmarks in

general, from a

...

marketing/sales standpoint saying "We don't believe in benchmarks and none are appropriate for our technology," does not go over very well with potential client/customers. (Nor do non-standard tests go over well since they don't allow "comparative shopping.")

My experience is that it is much better to at least subject yourself to such "standard benchmarks" and have a compelling "footnotes and explanations" to address any mismatch between your technology and an industry benchmark.

I would agree with Chris that benchmarks are generally biased but you are right they are essential for marketing purposes. So why not give them marketing non-sense ;-) I mean run medusa, just pure medusa no Zope to serve static html pages, and even better cache the pages in memory. It is Zope's web server after all.

Talking nonsense

Nothing nonsensical about it Pavlos, this is exactly the underhanded, rediculous types of methods other companies use to benchmark their products. In fact, this would probably be the *correct* way of benchmarking ZServer from a benchmark expert's POV. After all, to a person like that, Zope would be the backend and not really germain to the 'serving' nature of ZServer. Any other discussion along dynamic content lines goes straight into the "What is Dynamic? What is Content?" discussion. I'm sure ZD net and everybody else has an answer to that question, and I'm sure each answer is wildly different, will require a different benchmark, and all system will perform quite differently between them. Everybody's looking for the hard line in a world of soft planes. I'm also of the opinion that Zope and even other Open Source tools don't show up on these popular 'unbiased benchmarks' because we don't pay for advertising in those magazines. Quite simply, they don't know who we are because they are quite busy with the people they know and love so well, the one's who provide their paycheck. I don't think they are intentionally *excluding* products like Zope, but I have no doubt that there is an implicit preference to please the money. Hey, I please the money too. -Michel the comercialy bitter

Pavlos

_______________________________________________ Zope maillist - Zope@zope.org http://www.zope.org/mailman/listinfo/zope

(For developer-specific issues, use the companion list, zope-dev@zope.org - http://www.zope.org/mailman/listinfo/zope-dev )