Tim Larson wrote:
[Forwarding private email snippets (with permission) to list. More topics will follow under different subject lines.]
--- Marc Portier <[EMAIL PROTECTED]> wrote:
yep, and I'm afraid I haven't listened enough to the need you want to address, the more I think about the class-new stuff the more I end up thinking about some central catalog
--- Tim Larson <[EMAIL PROTECTED]> wrote:
There should be a central catalog, but I think we also need to be able to have classes that are local to a form. Now that the local case has been implemented someone can extend "new" to also be able to reference a central catalog.
BTW, I was thinking about allowing multiple catalogs and using "import" to reference in the relevant catalogs or specific classes from certain catalogs, similar to how import works in java source code. WDYT?
[Note this would also be similar to import/include in XSLT.]
--- Marc Portier <[EMAIL PROTECTED]> wrote:
I was thinking about augmenting the default-manager with a reference to a catalog where all the 'classes' are
and then have the new look those up from there, they should then have a logical-name, and should get a specific id in the new definition...
still thinking though
[End of email snippets]
More background from my side:
my bias towards central stuff is lowered, I see know how we can merge the two and indeed postpone the central thing until we have a more clear view on the advatages of these new things
however, my escape towards the central stuff was driven by a very strong (IMHO crucial) believe that all created definition and binding classes should become immutables: created once, reused all over, and that we should try to resolve all dynamics either during the build or inside the widget-instance-tree
in fact, the reason that I lowered my stress on central stuff, is that during discussions we somewhat found a conceptual way to support the local-class/new stuff while still holding onto the goal of immutables (although I need some more thought on that)
so to me, the roadmap: central catalog later makes sense
I would however like to here other visions on the immutables vision however...
regards, -marc= -- Marc Portier http://outerthought.org/ Outerthought - Open Source, Java & XML Competence Support Center Read my weblog at http://blogs.cocoondev.org/mpo/ [EMAIL PROTECTED] [EMAIL PROTECTED]
