Re: PARENT
Jean-Francis Balaguer (balaguer@artemedia.ch)
Thu, 22 May 1997 15:59:56 +0200
Stephan Diehl wrote:
>
> Sungwoo wrote:
> > 2)
> > Then, we should name a node explicitly when the node is
> > accessed by its child node ? Since the scene graph is
> > tree-structured, PARENT seems to cause no problem.
> > (I am not sure ...)
>
> Because of DEF/USE the scene graph is directed acyclic graph
> (in Script nodes there is also a way to produce a cycle,
> this was discussed in www-vrml a few months ago. This
> cycle is important to access the node containing a
> Script node from within the script).
>
> Here is how you get a DAG:
>
> DEF X SomeNode { }
>
> Transform { children [ USE X
> Transfrom { children USE X } ]
> }
>
> Now the inner and the outer Transform node are
> both parents of X !!!
>
> This doesn't mean that I don't want a PARENT keyword,
> I just cannot think of a clean semantics for it.
>
Well, as I mentioned in a previous mail, I solved this problem
in i3D using sharing and delegation. in i3D, multiple instances
are created with different objects that shares the value of
the VRML fields. Objects behaves as shared VRML nodes. Sharing also
make an optimal use of memory space. That way, we can answer context
sensitive queries like parent, global_transform.... since individual
objects maintains their own value for these fields....
--
_____________________________________________________________________________
Jean-Francis Balaguer
Technology Manager 3D/VRML
Artemedia Online SA e-mail: balaguer@artemedia.ch
rue de Morges 24 tel: +41 21-706-2084 fax: +41
21-706-2071
1023 Crissier - Switzerland www:
http://www.artemedia.ch/~balaguer/