[5188] in java-interest
Re: VRML
daemon@ATHENA.MIT.EDU (Pat Niemeyer)
Wed Jan 31 15:50:57 1996
Date: Tue, 30 Jan 1996 17:30:48 -0600 (CST)
From: Pat Niemeyer <pat@icon-stl.net>
To: Tony Healy <thealy@magna.com.au>
Cc: eknudson@vnet.IBM.COM, java-interest@java.Eng.Sun.COM
In-Reply-To: <199601280524.QAA10800@magna.com.au>
On Sun, 28 Jan 1996, Tony Healy wrote:
> 2. VRML browsers will not be written in java, because rendering speed is
> critical for them, and native code will always be faster than general
> purpose java code. That said, we will probably see some marketing of
> browsers purporting to be java, but which will still require native modules.
[disclaimer: I do not know what I'm talking about]
Java can be optimized as well, if not better than, C++. The only
excuse for there being no purely java based VRML browsers in the long term
is that they will want to take advantage of *hardware* rendering tools.
I think what is needed is a good 3D API. Then let us see who can produce
tools that are fast enough and in what languages.
> Existing 3D samples in java are clever but noticeably avoid large sizes and
> solid colors or shading, which require fast rendering.
Existing Java implementations are perhaps 20 times slower than C.
That will change soon...
Who cares about all those other instruction sets anyway?
Wait for the Java terminals ;)
Pat
-
This message was sent to the java-interest mailing list
Info: send 'help' to java-interest-request@java.sun.com
