Re: [Gnu3dkit-discuss] considering G3DGeometry

[Brent Gulanowski]

On Sunday, March 16, 2003, at 06:41  PM, Philippe C.D. Robert wrote:

> > Please elaborate. I have been thinking about this a lot and its seems 
> > worthy of investigation, but if you know any serious reasons that 
> > would make it unrealistic, I would appreciate if you could mention 
> > them. If I want to browse
>
> If you create a new geometry class how do you make sure that every 
> existing renderer can handle it? If you write let's say a 3D modeller, 
> you would maybe want to use parametric surfaces together w/ an OGL 
> renderer, but how would you turn the created surfaces let's say into 
> equivalent implicit surfaces used by another, raytracing based > backend?
>
> There are many questions like these which are not easy to answer.

Undoubtedly! These questions are, however, more in line with my skills 
and interests than, e.g.: implementing shader parsers and designing 
shaders. If it comes down to a mathematical solution, again I'll 
probably leave it up to someone else, but as for setting the priorities 
of the representation types, and synchronizing them, and writing the 
state management logic which determines which ones to use when, this 
intrigues me very much.

I don't think you can make sure that every renderer instance can have a 
visually identical (or even very similar) representation. It would 
perhaps be possible to generate an approximation, or use a default rep. 
Given a PDF, how does NSImage produce a TIFF or a JPEG or whatever? 
Math of some sort! It must involve some kind of dependency graph from 
one representation which stores the most overall information down to 
that which stores the least. Potentially with two manually produced 
reps (or one manual, the other generated and tweaked by user/artist), 
you might then have enough fundamental information to generate most 
other possible reps, given a library of translation algorithms and a 
support class to implement them and return the desired rep.

I have a lot more thinking to do. Sometimes the applications of 3D tech 
that I am interested in would use 3D objects merely as a means of 
communication. This is very different than using 3D to render 
photo-realistic images. I know that games are not your focus (maybe you 
don't even like them), but they are another example where 
representation has symbolic, not literal, purpose, no matter the 
current obsession with realism in 3D games. I am interested in 3D 
virtual toys, in 3D user interfaces, and the use of 3D for 
visualization of information that is not inherently graphical (like 
structural relationships).

In such applications, representations are much more fluid, and much 
more dependent upon the viewing environment. Like with a 2D graphical 
user interface, what is a cube in one place might look like a flower or 
a tree or a television or anything else in other places. For me, visual 
representation is merely a surface applied onto a structural system. A 
scene graph is an example of a very common data organization method, 
which can be used for much more than strictly spatial data. I am highly 
interested in finding structural similarities between 3D scenes and 
other information matrices (not algebra!).

Brent Gulanowski
--
Mac game development news and discussion
http://www.idevgames.com