RE: [cgmo-webcgm] XCF and "inherit" value

[Dieter Weidenbruck <dieter@itedo.com>]

Alternative:
The root node could also be the picture, not the metafile.
Reason:
In a multi-picture CGM, each picture is independent from the others.
So the metafile node may be considered as the root for the picture
list, however, for any APS within an independent picture the picture
node should be the root.

Pros and Cons?


> -----Original Message-----
> From: Lofton Henderson [mailto:lofton@rockynet.com]
> Sent: Tuesday, May 24, 2005 4:16 PM
> To: dieter@itedo.com; cgmo-webcgm@lists.oasis-open.org
> Subject: RE: [cgmo-webcgm] XCF and "inherit" value
>
>
> At 10:00 AM 5/24/2005 +0200, =?us-ascii?Q?Dieter__Weidenbruck?= wrote:
> >[...]
> > > It seems to me that the parentNode attribute on a layer or a
> > > top-level APS
> > > would be the Picture, right?  So for the purposes of the
> > > inheritance model
> > > discussion, do we need to say that it is a "virtual APS" that has the
> > > Initial Value of the ApsAttrs in question?  (And if "yes",
> then ... well
> > > the Picture isn't the root yet, is it?  The Picture's parentNode
> > > attribute
> > > is the Metafile, right?  Same question there.)
> > >
> > > Thoughts?  How can we deal with this cleanly?
> >the "root" should be the metafile IMO, and it needs to be treated as a
> >virtual
> >APS as far as visibility, interactivity, and style properties
> are concerned.
>
> That's clean, and along the lines I was thinking.  I think a short
> paragraph somewhere in 5.4 should take care of it, and it
> probably doesn't
> need to appear anywhere else.
>
> Btw, I reread 5.4 fairly carefully, hoping that it was already
> covered.  But I don't think it is, quite.  At least not clearly and
> unambiguously.  Values can inherit to and through nodes where
> they are not
> applicable.  But it needs to be clarified that steps #1-3 in
> 5.4.1.1 should
> be applied to the Picture and Metafile as if they were APSs, in order to
> resolve the SV and the CV of the top-level APSs (layer, grobject, etc) in
> the picture.
>
> If we solve it in 5.4, the oddity about "inherit" value for 'layer' goes
> away -- not a problem.
>
> -Lofton.
>
>
>