RE: Why does DEF do instancing?

Henry Nash ([email protected])
Fri, 13 Oct 1995 10:15:42 -0700


I agree, it does seem conceptually wrong. Also, although, won't the =
workaround cause the VRML client to actually create sphere (and in =
theory render it). Do you have to put it out of sight?!?! =20

What if this was a complex object like some new kind of video stream - =
having defined version around that we didn't want could become a =
problem. =20

Is there any support for either changing DEF (Oh, I hope not, too =
dangerous) or supporting a new definition keyword (maybe DECLARE ?) that =
would actually declare the "type" but not create an instance of it. =
Does anyone know if Inventor supports this ? (I took a quick peek, but =
couldn't see anything).

Henry

----------
From: Vassilis Bourdakis[SMTP:[email protected]]
Sent: Friday, October 13, 1995 8:25 AM
To: VRML-Mailing List
Subject: Re: Why does DEF do instancing?

Thomas Bornfleth wrote:
>=20
> But it isn't hard to do a workaround.
> Just put another Separator around the DEF node, in which you do a =
translation.
> So your example would look like:
>=20
> Separator {
> Transform { ... somewhere }
> DEF RedSphere Separator {
> Material {
> emissiveColor 1 0 0
> }
> sphere { }
> }
> }
>=20

The problem is that IMHO it is conseptually wrong! When you define
something you simply define it as far as I'm concerned. That is the
way autocad does it and other CAD s/w as well and it does make
sense.

It is also easier to deal with. I'm currently writing an ACAD to
IV or VRML translator and came across this error trying to
deal with blocks =3D=3D DEFs and insertions =3D=3D USE. Took some time =
to
figure that out...

Cheer

--=20
--- --- ---- ---- ----=20
Dr. Vassilis Bourdakis ---- ---- ----- -----=20
Tel +44 (0)1225 826475 CASA, Bath University =20
[email protected] http://fos.bath.ac.uk


  • Next message: Joseph Lombardo: "Re: Q#3 Crystal Balls or something...."
  • Previous message: Sean O. Scott: "Is there a vrml viewer for Power Macs yet???"