On Mon, 2002-04-08 at 20:41, Reed Hedges wrote:
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
>
> On Monday, April 8, 2002, at 08:14 PM, Braden McDaniel wrote:
>
> >
> >> Eventually I want to support Script
> >> nodes (which would affect my object structure).
> >
> > Aha... Now this is a whole 'nuther ball of wax. What kind of Script node
> > support are you looking for? If you mean *complete* Script node support,
> > then you need a whole VRML runtime engine, and OpenVRML definitely fits
> > the bill there.
>
> Yeah, pretty complete. ROUTEs too. What is the interface to Routes and
> Script nodes? I ran doxygen (by hand, after modifyiing Doxyfile.in --
> configure could not find my GL libs and aborted) but only noticed the
> abstract ScriptObject parent class...
Check out the rearch branch. Tag is OpenVRML-REARCH_BRANCH. The doxygen
docs there are *much* better.
> > JavaScript is adequate, but incomplete. Check the task list. Java is
> > there, but nascent.
>
> Yeah, JavaScript.
>
> >> Can I build OpenVRML without any OpenGL rendering features and just use
> >> it's parsing and scripting features?
> >
> > Yes.
>
> An option to configure?
No, but we need one. It's not hard; I'll do it before our next release.
(After the rearch branch merges back with the HEAD.)
> > It would be interesting to see a bridge between OpenVRML and some other
> > 3D runtime system. Such a project would definitely be helpful in
> > discovering what needs to be exposed in our API.
>
> I won't have much time for this until the summer, but I'll be working on
> it heavily then. The VOS 3D stuff is informed by VRML, but not really
> based on it at all... so it will be a bit irritating bridging them :).
> (It's not a scene graph per se, though the transformations are
> hierarchical; we only have a few primatives and material
> characteristics, but will be adding more. And of course, it's always
> the little things, right?)
The goal of the rearchitecture is to separate node implementations from
the core. We're not there yet; but the rearchitecture has made the
points of cohesion clear, and they will be eliminated one by one. I
mention this because it sounds like it will have significant
implications for the design of such a bridge.
> Also, I primarily work in Mac OSX. Any work with OpenVRML on that
> platform yet? (One irritating thing is the weird places it puts some
> libraries, such as OpenGL and GLUT :) It's really easy to "port" stuff
> from Unix-ish platforms though, so I can do that pretty easily.
It was at least Mostly Working on OS X at some point. OS X is definitely
someplace we want to work well.
--
Braden McDaniel e-mail: <br...@en...>
<http://endoframe.com> Jabber: <br...@ja...>
|
