|
From: | Gerard Iglesias |
Subject: | Re: [Gnu3dkit-dev] OGL 2.0 shading |
Date: | Fri, 31 Jan 2003 09:59:42 +0100 |
User-agent: | Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.0.0) Gecko/20020530 |
Philippe C.D. Robert wrote:
Yes, exactly. Either you use a node transformation to position the matrix and/or use a so called rotator to orientate the cam.
I agree, it is OK for me
My idea was to provide a clean MVC concept here, so the camera is the controller which displays the model (=scene) using a view. This is an important difference to the original NeXT 3DKit, where the camera was a subclass of NSView.
Your point is good...
But it would be better to me to separate the view functionalities to the frustum ones.
What do you mean exactly here? Should we again provide a separate G3DFrustum object as in 0.3.x? I removed this separation because it did IMHO not add value to the design but cause communication overhead during rendering (when view-frustum culling).
The overhead is so critical, What I think about that, is the capability for the controller to change the viewing setting from a predefined set, and I don't see other things than an object or a C struct (if the overhead is critical)
On an other subject, did you see that NVidia is porting Cg to MacOSX (I read a post on the mac-opengl maillist)
Regards
[Prev in Thread] | Current Thread | [Next in Thread] |