Re: [eclipse-incubator-e4-dev] YAMI (Yet Another Model Interface)

[Kevin McGuire <Kevin_McGuire@xxxxxxxxxx>]

> > Unfortunately, if used directly from Java code, an API like this
> > does not lead to very readable code.  The only defense I
have 
> > against this is that I don't think you would want to program
against
> > this API directly.  It would make a nice basis for data
binding 
> > though - who says you have to program the API directly?
> 
> I think this is a very cool idea.  I could see how this could
be 
> applied very nicely against an underlying EMF model to induce a 
> virtual structure for the purposes of viewing in a way that's much
> more flexible than the more rigid content providers.  It could
be 
> used to view IConfigurationElement and IMemento instances as well.
 
> I don't think it's intended as a way to unify general purpose data
> manipulation for these two things.  In that way, I don't consider
it
> a YAMI. :-P  Rather it seems to be a way to unify all the JFace
> content providers as well as avoiding a proliferation of additional
> ones in the future.  Am I interpretting this correctly?

That was my interpretation as well. Its a shallow
but consistent model for the UI viewer, just as content providers are,
but at a higher level.  I found it handy for building the web demo
we showed at the BOF.  It does however lead to ambiguities around
level of modelling (do I model the entire perspective, or just the trees/lists?
 what about the connections between them?) and around reuse (if I
have two navigators that are very similar how do I reuse the observable
or are they simple enough to write that writing two isn't a big deal?).
 In the latter case they are trivial to write except of course the
delta processing is hard work and that part you would like to reuse.

Kevin