[Zope3-dev] Twisted Publisher and Zope 2

[Jim Fulton]

Sidnei da Silva wrote:
> On Mon, Dec 12, 2005 at 06:25:53PM -0500, Jim Fulton wrote:
> | Sidnei da Silva wrote:
> | ...
> | >Sounds good to me. By quickly looking Zope 3's requests have mostly
> | >the same methods and features from Zope2's. However sems like most
> | >methods were renamed for consistency (eg: supports_retry ->
> | >supportsRetry).
> | 
> | There are a number of things I can think of off the top of my head:
> | 
> | - getting request-based URLs.  For example request/URL/1 vs request/URL1
> 
> Uh, Zope 3 has the first I guess?

Yup.  These should be easy enough to combine.

...

> | >The greatest lacking functionality in Zope 3 seems to be the lack of a
> | >'lazy' namespace, which is used primariliy for the 'SESSION' object in
> | >Zope 2. How do people feel about adding that to Zope 3?
> | 
> | I'm not familar with this. Where is it documented?
> 
> Here's what is in the docstring for HTTPRequest:
> 
>       - Lazy Data
> 
>         These are callables which are deferred until explicitly
>         referenced, at which point they are resolved and stored as
>         application data.
> 
> Haven't found much else documentation.

Are these basically lazy computed request keys/attributes then?

> There's a 'set_lazy' method in HTTPRequest, and that's what the
> session machinery uses to bind 'getSessionData' as 'SESSION'. This
> specific case sucks though, because as 'SESSION' appears when doing
> request.keys() is pretty common to create sessions implicitly by
> iterating through the request.

Yeah, explicit is better than implicit.

I don't want to add this to Z3, but it should be included in the
new combined z2/z3 request.  We really need to take a look at the
combined API and decide what we want in the long run.

Jim

-- 
Jim Fulton           mailto:jim at zope.com       Python Powered!
CTO                  (540) 361-1714            http://www.python.org
Zope Corporation     http://www.zope.com       http://www.zope.org