in defense of architecture

[Iain Duncan]

Lately ( as a side effect of growth I expect ), I've noticed another wave of emails on the lists bemoaning the inclusion of the ZCA, zcml, zodb and traversal in pyramid. I suppose it's natural that people don't read defense-of-architecture docs when first coming to an framework, but it's also natural that people like Chris will get sick of defending their choices over and over again. ( kudos for being as polite as you have been! )

I'm wondering, does anyone have any good ideas on how we can improve this? I think Chris et. al. have done a great job of making it all optional, and invisible if you don't want to use it, but it seems to me that if possible, maybe more could be done to explain why these are all very useful parts of the framework for users looking at pyramid as an option. Would more example pages be useful? more blog posts? I have been tossing around starting a python blog to write more about our experiences. I don't know whether that would be useful, or more example apps, or what.

thanks

Iain

[Graham Higgins]

There's this, from a couple of years ago, back when the nascent Pyramid was going by the pseudonym "Marco". I encountered the PCA (as the ZCA should have been called) and went hunting around for some other approaches to providing a component architecture:

http://bel-epa.com/notes/Marco/notes.xml

I didn't spot the Qt approach that you mentioned in your other post but I would be interested in a comparative review.

The "Zope shudders" are simply ill-informed prejudice - I've yet to see a substantial argument to be made against using the ZCA in a well-defined role as a component architecture specification. If folks want to diss the ZCA, they should be proposing alternative approaches and making cogent arguments in support of them. Actually, I'd /really/ be interested in any informed discussion of component architecture issues that used Pyramid as a springboard.

Cheers,

Graham Higgins.