Newsgroups: rec.autos.simulators
Path: news.jprc.com!dca1-feed2.news.digex.net!digex!news1.radix.net!tor-nx1.netcom.ca!ix.netcom.com!not-for-mail
From: dripton@netcom.com (David Ripton)
Subject: Re: GPL Framerate Hit Caused by AI, not graphics
Message-ID: <driptonF07I78.9yv@netcom.com>
Sender: dripton@netcom5.netcom.com
Reply-To: dripton@netcom.com
X-Newsreader: trn 4.0-test67 (15 July 1998)
References: <6u5oti$nce$1@news.interlog.com> <3612389A.B136424@helene.uio.no> <6uvd7r$p18$1@nnrp1.dejanews.com> <3614CAE0.F9031FEE@helene.uio.no>
Date: Fri, 2 Oct 1998 15:40:20 GMT
Lines: 27
Xref: news.jprc.com rec.autos.simulators:45570

In article <3614CAE0.F9031FEE@helene.uio.no>,
asgeir nesen <asgeirn@helene.uio.no> wrote:
>Voodoo II is a marvellous product, with a very specific customer target, namely the
>top-end users... It is not the best product in all contexts, as you rightfully agree

You keep arguing this, but that doesn't make it correct.

Go look at some comparative 3D card benchmarks.  (There's a big 
3D card comparison on www.tomshardware.com, for example, which
uses several different benchmarks with several different CPUs with
several different cards.)  Voodoo and Voodoo 2 cards do very well 
on lower-end systems in most games, and scale less with CPU than 
most other 3D cards in most games.  Papyrus games are an unusual 
case, as they were originally written for Rendition's strengths 
(like unified video memory) and weaknesses (like no Z-buffer) and 
then ported to 3Dfx.  You just can't make general statements about 
3D card CPU scalability from one engine's results.

The statement you make about the Voodoo 2 is much more applicable
to, say, the TNT.  Crazy fast on high-end systems; not a very
good choice on low-end systems.  (And unusable in GPL, sigh.
I hope Papyrus sees the light and makes an OpenGL or Direct3D
version of their engine.)

-- 
David Ripton    dripton@netcom.com
spamgard(tm): To email me, put "geek" in your Subject line.
