|
|
|
|
|
Re: Nested Notification Question [message #428804 is a reply to message #428802] |
Tue, 31 March 2009 00:21 |
Ed Merks Messages: 33133 Registered: July 2009 |
Senior Member |
|
|
Steve,
Comments below.
Steven Punte wrote:
> Ed Merks wrote:
>
>>> I only have the one adapter at the very top of my emf tree structure
>>> which calls refresh of the treeviewer.
>> You're not using the generated item providers?
>
> No, I am not. They are excellent at portraying the exact tree from a
> software developer perspective, but from an end-user perspective they
> are not sufficiently user friendly.
They determine which features shown as children and how the label of
each item is surfaced. They're infinitely malleable. I can't imagine
how you'd induce an end-user tree without the same concepts applying.
Indeed that seems the very problem with which you are faced...
>
>>> Do I need to create and apply an adapter/observer at EVERY parent
>>> node of my emf tree, or is there a simpler more elegant way?
>> Item providers are very simple and elegant. I have a feeling you've
>> overlooked them.
>
> I am certainly in the dark with respect to the machinery that manages
> these providers.
So in that darkness you seek to find an alternative solution to the very
same problem.... Likely a simple one that has yet to occur to anyone
else...
> Does EMF.edit machinery instantiate an ItemProvider for each node in
> the users EMF tree?
You'll note there is the concept of singletons. I.e., stateless item
providers, the default, which, given an instance object, can answer all
the questions about how that object should be viewed as a tree node.
> And does it create and destroy such ItemProviders as the user add and
> deletes nodes?
Why ask when a fully functional editor is generated for your detailed
analysis? The debugger will answer every question more quickly than I.
>
> Thanks for you help.
> Steve Punte
>
>
Ed Merks
Professional Support: https://www.macromodeling.com/
|
|
|
Powered by
FUDForum. Page generated in 0.01749 seconds