[Zope3-dev] Re: uniformly configurable request factories: bugfix or proposal?

[Stephan Richter]

On Thursday 04 August 2005 23:34, Tres Seaver wrote:
> > Shrug, this is a risky change. If something goes wrong there, a lot of
> > code will be screwed. I agree it is fairly trivial to fix, but I would
> > prefer it to be done right. I plan to address the entire
> > HTTPRequest/Publication/... issue with the Twisted integration work.
>
> Can you point me to the discussion of the tradeoffs involved in moving
> to Twisted?

I have not written the proposal yet, but you can play around with the branch 
if you like.

> I have a (perhaps unfounded) sense that its HTTP server 
> implementation has serious scaling problems, even compared to ZServer.

Not the new web2 server. Itamar and I did some very quick (non-scientific) 
profiling and web2 was just slightly slower than zserver. It was just a very 
small fraction in comparison to the time the publisher requires.

> Perhaps that has changed over time, but at the time I acquired the
> opinion, it seemed to be the sense of the Twisted community that HTTP
> wasn't "shiny" enough to get lots of attention, in the face of "sexy"
> work to add new features.

web2 is a reimplementation from James Knight that concentrates more on the 
HTTP standard itself and provides a solid WSGI interface.

> Does anybody have experience running Twisted at "enterprise" scale?  By
> that, I mean:
>
>   - pagers go off when it falls over;
>
>   - you know your "committed information rate", and watch your bursts
>     carefully;
>
> etc.  I know that the feature set which Twisted makes available is
> interesting, but I would be *strongly* opposed trading away stability
> and scale for HTTP in exchange for yet-another-protocol-which-noone-
> actually uses (likely an unfair characterization, I admit).

Jim already convinced me to leave zserver has an alternative. Not quiet sure 
how I will do this, but I hope it will be as easy as switching a few options 
in zope.conf.

> I am *not* opposed to the idea of making Zope more friendly to running
> behind Twisted;  I *do* think we need to consider carefully whether we
> will be wise to make Twisted the *only* way to run Zope3 for HTTP.

It will not be the only way, but the default one.

Regards,
Stephan
-- 
Stephan Richter
CBU Physics & Chemistry (B.S.) / Tufts Physics (Ph.D. student)
Web2k - Web Software Design, Development and Training