Xref: utzoo comp.sys.amiga:74918 comp.sys.amiga.tech:17149 comp.sys.amiga.hardware:5204
Path: utzoo!utgpu!news-server.csri.toronto.edu!dgp.toronto.edu!kevins
Newsgroups: comp.sys.amiga,comp.sys.amiga.tech,comp.sys.amiga.hardware
From: kevins@dgp.toronto.edu (Kevin Schlueter)
Subject: Re: Retargetable graphics. Temporary Stop-gaps.
Message-ID: <1990Dec20.164538.24027@jarvis.csri.toronto.edu>
Keywords: rtg,iff,24bit,video hardware
Organization: CSRI, University of Toronto
References: <12437@life.ai.mit.edu> <16612@cbmvax.commodore.com> <7307@sugar.hackercorp.com>
Date: 20 Dec 90 21:45:38 GMT
Lines: 19

Some people are asking why we should accept the limitation of not being
able to run intuition on new displays.  Basically, the issue is that
of 2D graphics vs 3D graphics.  By 2D graphics, I'm referring to the
graphics used for window and gadget imagery and the graphics found in
draw programs, basic paint programs and other common apps.
These programs require  bitblt and
its relatives, rectangle/polygon draw/fill, flood fill, pattern fill,
optimized text display routines, etc).  By 3D graphics, I'm talking about
the kind of graphics required by 3-D modelling and animation programs --
i.e. primitives to support 3D affine transformations, complicated clipping,
various types of shading/illumination models, and texture mapping.  

Often, when designing a framebuffer, you make tradeoffs that favour one
type of graphics over the other, in order to keep costs down.  For this
reason, we should consider whether we need to run _all_ applications on
fancy 24 bits frame buffers or whether we should make 24 bit displays that
are optimized for modelling and animation.

Just a thought...