...with a dynamically-generated schema since the set of legal constructs depends on what is on the classpath...
Or something like that.
;-)
On Mon, Aug 11, 2008 at 8:43 AM, Ed Merks
<ed.merks@xxxxxxxxx> wrote:
Kevin,
It's almost like you need a model of the widgets. :-P
Kevin McGuire wrote:
Hi Ralf,
> Yes, that's also our preferred approach. In
RAP
we have an extension
> point for themeable widgets anyway (the RAP theming has to know
about
> themeable widgets to keep styling and layout in sync between
client
and
> server). This is also a suitable place to register such an element
name.
> I'm not sure whether such an extension point would also be
appropriate
> for Eclipse in general. Who would have to register with such an
> extension point? The component provider, e.g. Nebula?
Agree, I think that's along the
lines
of how it would work. Namespaces aside, for each extended widget,
someone (presumably the widget component provider) needs to :
- publish the set of properties
each widget
supports so that a developer knows what to write in the curly CSS
brackets
(and we may want for a validator),
- describe to the platform the
class of
the widget which will match the name appearing in the CSS file,
- describe to the platform the
methods
to invoke matching those property values.
Kevin
_______________________________________________
eclipse-incubator-e4-dev mailing list
eclipse-incubator-e4-dev@xxxxxxxxxxx
https://dev.eclipse.org/mailman/listinfo/eclipse-incubator-e4-dev
