Yes. Then I can write my own code for a Templar marching around the Sun Dome Temple, a Templar marching between two fixed points, a Templar standing still and yawning occasionally, a Templar having a drink in a bar (water, of course!), and so on. All from one model. With code to detect when the viewer is near him so he can react.
I'd be interested in seeing how Poser exports animations to VRML: it may well do the donkey-work for me, and just need the code tweaking a bit. (An excuse to write Javascript; goodie!)
> I can get the basic model
> mesh into Lightwave OBJ, as 3ds, as anim8or (a new format with some
> neat features but not debugged) or even as a C source file. OBJ,
3ds,
> and C probably won't have joint parameters. I'll dig around for
more
> information on this. Unfortunately, I have no VRML authoring
software
The best VRML authoring SW is Notepad, or some similar text editor, so I suspect you do.
> Where is there a good definition of the VRML model format?
http://hiwaay.net/~crispen/vrml/
is a good VRML reference site, and will have the up-to-date links to the language spec.
>
> > (What I'd really like is a Sun Dome Templar, with limbs done as
> > separate transforms so I can animate them, and with a bitmapped
face
> Do you need the limbs to be done
> as distinct *objects*?
Yes. With positional relationships defined between them, (aka "transforms" to me: I have no idea what Poser calls them). So the arm is defined as having its rotational centre at the shoulder joint, and its position such that the top touchs the shoulder bit of the body with a bit of overlap. I can do this myself, given the objects: it's defining complex shapes where Notepad starts to come apart as an editor.
The face would need a nice simple bit-map, with the texture-transform coords set up to define how the image fits on to the model: then I just swap bit-maps as required. Rough edges would be hidden by the helmet in any case.
> However, as I've said, VRML is out there on the far fringe, so I'm
not
> much up on it.
I'm not aware of many other formats that give small enough file sizes to be useful on the Web. The models John Hughes has posted, frex, are HUGE and take forever to download a single figure. Beautiful, but put twenty of those in a city and start moving them around, down a 56K modem, and you're in trouble. (In fact, put one of them on-line, static, and it still takes ages).
> essentially, items where the final
> product is actually a "render" of the model and transformations.)
And there's the difference. I have no interest in a "render": I want a fully interactive city, on-line. So for a non-textured human model, a file size of under 10K is a must.
> If you had Poser, this would be a trivial matter, as would any preset
> animation.
I could get it, no doubt: but would this give me something I could use on-line, or just something I could look at locally?
Powered by hypermail