>...Also, how am I supposed to know where all of my models will end up and all
>the possible views that will be needed?
Well, nobody says you have to. As an object designer, you have the ability
to bundle interfaces to some useful coordinate systems with your object,
and can provide whichever ones you feel like supporting (hopefully at least
the "standard" one, once we all agree on it :) ). As an object user, if
the object comes ready with the coordinate system you want to use, great!
Otherwise, you have use a standard conversion, provide it yourself, or use
someone else's object.
>>1. A "furniture view"
andrieu@alta.com:
>Why is this useful for quickly placing objects against anything other than
>the back wall?
>>2. An "icon view"
>But why would I want to use this view in anything other than a catalog of
>VRML objects?
You wouldn't: these are special-case coordinate systems that are
nonetheless useful enough that they might be worth supporting. (Pictures
are almost always hung on a wall, right?) Or not. Maybe other coordinate
systems will be popular.
Okay, a couple more (admittedly over-dramatic) example to try to convince
people that multiple coordinate systems might be useful, before I let this
die...
Suppose it's 2010, there's millions of objects out there, and we all
have parallel PC's and gigabit-per-second networks, and someone wants to
put together a virtual Earth. One neat way to do it would be to define a
new "earth-coordinate-system", with origin at the center of the earth, z
towards the North Pole, etc, and ask everyone who wants to put something on
it to add a matrix to their object describing its location. But, they
wouldn't want to give up their "standard-coordinate-system" matrix to do
that (especially since most people's buildings would end up tilted 30-60
degrees), and no single person would want to have to take all these objects
and place each of them by hand at the spot on the Earth where it belongs.
Similiarly, what if someone decides to build a "virtual Tilvertonia", by
virtue of the Tilvertonia TV show being a major hit, and some people want
to put their objects in "Tilvertonia-coordinate-system" as well as in
"earth-coordinate-system and "standard-coordinate-system"?
The idea here is not to limit people in where they're allowed to put
objects, or to save FPU cycles (as a few people seemed to think), but to
save object-placement time. I'm assuming (not having done it myself) that,
no matter your choice of user interface, it's somewhat tedious to have to
convert to the right coordinate system, rotate, scale, and precisely line
up hundreds or perhaps thousands of objects to create a scene. So, in the
admittedly limited cases where the object designers can help out with the
boring stuff, it would be a win to give them a mechanism to do so.
(That being said, I'll defer to the judgement of those more experienced in
computer graphics than I. I'm subscribed only to the digest list, so
please cc: responses directly. Thanks.)
| Brian Slesinsky | bslesins@netcom.com | bslesins@us.oracle.com (work) |