Define the properties of the model. Then do the
XML DTDs. Objects first: format last.
Look carefully at the MS site (love 'em
or loathe 'em, they have the best browser support
for XML). There could be potentials if X3D
tag sets for 3D can use data sources (eg, use
data islands).
To define a DTD for storytelling, design a
hierarchical set of properties for a story.
Tag it last. There are DTDs out there
from the SGML days for theatre scripts but
that is linear unless you provide controls.
Again, the idea of feedback-mediated hypermedia
is based on common events for a defined set of
controls. How do you work controls into a
story? Can you get beyond a game doing that?
Do you have to? Dynamic node navigation for a
story in an *environment* may only mean the
characters can control you. As in life, you
can ignore any participant in a conversation,
and you can be ignored, but that is solitaire
with blank cards.
To do this, I would create a set of
DTDs, one for each modular piece of the
storytelling environment (eg, H-Anim characters,
scene characters (environments are characters
that speak their own language and you have to learn it),
in other words, design the active objects.
If you need a script DTD,
all it is doing is grouping these and declaring
their relationships. Then I would combine these
into aggregates using namespaces.
Len Bullard
Intergraph Public Safety
[EMAIL PROTECTED]
Ekam sat.h, Vipraah bahudhaa vadanti.
Daamyata. Datta. Dayadhvam.h
> -----Original Message-----
> From: Michael St. Hippolyte [SMTP:[EMAIL PROTECTED]]
>
> > I'm also very concerned with separating the structure from the
> >presentation (layout) which sort of implies XML. Someone maybe have seen
> >a DTD for this (len:)?
>
> If no one knows of a good one, I suggest we make one up ourselves. After
> all, at the rate DTD's are popping up, pretty soon they'll be like home
> pages -- everyone will have one.
>
> > The visualization is of course done with VRML :)
>
> Generated or hand-crafted? (Just curious, I love them both...)
>
>
