Hash: SHA1

Jim Fulton wrote:
> Stephan Richter wrote:


>> I agree with your assessment. It is extremely difficult to figure out which 
>> WSGI server fulfills Zope's criteria. In fact, I would suspect that only 
>> ZServer (Zope 2 and 3 version) does, because noone else has such strong 
>> requirements.
> What requirements?  If we have such requirements, I suggest we reevaluate 
> them.
> *We do not want to be in the server business!*

We have performance and reliability expectations which come from running
mission-critical applications.  Lots of the rest of the folks interested
in servers don't have those requirements (yet, anyway), and hence aren't
motivated to address them in their externally-maintained server

What is *worse* than maintaining our own server is trying to track
development on somebody else's, where their goals don't match ours.  I'm
guessing that the amount of effort required to "maintain" the Zope2
server per month is less than the aggregate time spent by the community
reading this thread. ;)

> Perhaps you are thinking of the buffering requirement.  I would suggest that
> this be part of the benchmark.  That is, there should be a test of whether 
> someone
> can DOS a server by connecting with a non-zero content length and then not 
> providing
> input.
>> I would love to see such a profiling tool too, not only for 
>> testing servers, but applications as well.
> Applications is totally out of scope for this IMO.  Everybody
> wants benchmarks that apply across applications, but that is a pipe dream.
> OTOH, it should be feasible to come up with a reasonable set of tests that
> can be used to compare servers.

- --
Tres Seaver          +1 540-429-0999          [EMAIL PROTECTED]
Palladion Software   "Excellence by Design"    http://palladion.com
Version: GnuPG v1.4.2.2 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org


Zope3-dev mailing list
Unsub: http://mail.zope.org/mailman/options/zope3-dev/archive%40mail-archive.com

Reply via email to