Re: RFC: Schema Refactoring, first steps

On Fri, 14 Jul 2006 02:48:59 +0100, Martin Aspeli wrote:

> Rob Miller wrote:
>> On Thu, 13 Jul 2006 15:24:55 -0400, whit wrote:
>> 
>>> I added rocky's requested clarifications. feel free to add language (or
>>> proprosals):
>>>
>>> https://svn.plone.org/svn/archetypes/proposal/schemaproprosal.txt
>> 
>> i've read all of this, and, while i agree in general with the design
>> sentiments, i'm not quite convinced that it's the right way to go.
>> honestly, my thought is that i'm not sure we should be investing this
>> much energy in making AT better.  as you all know from my other posts,
>> i'm strongly in support of making the Z3 stuff and the AT stuff both
>> play together well within the Plone UI and within the same content
>> objects. but that's specifically so that we can ease the transition off
>> of AT.  i'm quite content to change AT just enough so that we can make
>> this happen smoothly, and i'm suspicious of any major efforts that go
>> toward improving AT rather than toward improving the Z3 story.
> 
> Whilst I agree with this, I think we need to be careful about the
> proverbial "stop energy". I don't think that's what you're saying here,
> but if we say "we can't do any innovation in Archetypes", we are cutting
> off a valuable source of use cases for future work, and we are choking off
> innovation in the framework that still supports 90% of the solution out
> there (and on which core Plone depends heavily).

i agree 100%.  easy wins that don't require a lot of effort, great.  more
than that seems not worth the effort.

> The better-interoperbility argument you make is really good (though I
> can't envisage how it'd work right now),

i imagine having the 'main' slot of main_template.pt invoke a viewlet
container.  then we'd write viewlets that can render AT schema widgets,
and other viewlets that can render Z3 widgets.  perhaps we can embed some
display organizational info that the viewlet manager would use as Z3
annotations on the FTI for a type, or even on individual instances.

trickier would be handling the validation, since the AT validation is
usually driven by CMFFormController while the Z3 validation is driven by
formlib.  it may be possible to invoke these processes one after the
other... i'd have to play around to see if that were really possible.

> and probably what will allow us
> to strike that balance in the end. All I'm saying is, let's be pragmatic
> about what we can achieve with AT in the short term, to the greatest
> cost-benefit, as well as where want to go in the future.

definitely.

-r




-------------------------------------------------------------------------
Using Tomcat but need to do more? Need to support web services, security?
Get stuff done quickly with pre-integrated technology to make your job easier
Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642