Re: Rdf correct use ? (anonResources everywhere ?)

Resume : I want to follow custom predicate in the RDF to build Trees, my 
example is to follow "subClassOf" predicate to build the tree class of 
my objects.

> I believe that you can persuade the builder to follow custom containment 
> properties but I don't know the syntax nor if it would apply to your 
> particular RDF graph.
> 

Well, I'm now quite aware of these rules and the template building 
system (I just read all and tryed a lot of thing after building various 
simili-solutions in slow and overbloated javascript).

These are my constations :

- template system is awesome... in theory : you can really build 
anything you want from any datasource.

- practically, it is really hard to master (no debug output) and seems 
limited !, and I could'nt find a "n.p.m.template" newsgroup :( to ask 
questions.

I couldn't manage drawing a correct tree following a "subClassOf" 
property (which defines a tree by its leafs). Well I actually managed to 
draw such a tree but the containment property that is used to draw the 
"+" to expand a branch is displayed everywhere (even for non-branch 
element) because the template builder uses ds.ArcLabelsOut() to find if 
our resource has our property linking _OUT_ from it. In my case : 
subClassOf is in ALL classes (as all class is subClassOf at least 
sub-class of "Resource"), this explain all the element displayed have 
the "+" to expand there branch.

I would need to "reverse" the way it was thought to gain an extra inch 
of modularity and build these tree correctly. We would need to give the 
opportunity to look if there's no property linking _IN_ (and not out) 
with (for example) the ds.ArcLabelsIn(). What I want to do is to take 
the containment property the other way round...

As anyone an idea ? a solution ? a hint ? hacking the tree.xml would be 
of any help or you think it is in the template cpp code that relies 
solution of all this ?

Thanks in advance
Valentin