Article View: sci.virtual-worlds
Article #3094Re: TECH: Arbitrary input devices
From: Bob Pendleton
Date: Wed, 06 Jan 1993 22:44
Date: Wed, 06 Jan 1993 22:44
44 lines
2230 bytes
2230 bytes
>From article <1ib69aINNr9b@shelley.u.washington.edu>, by thinman@netcom.com (Te
chnically Sweet):
>
>
> X is a mess and structurally all wrong for VR work.
Hmm...
But, just about every high performance 3d graphics system is embedded in X.
> My system uses this technique: data is fed to X number of input filter
> objects. When a particular object decides that input has happened, it
> sends an "Input Happened" message to its interested parties with
> itself as the argument.
This is pretty much the same way X events work. The rest of the method
you describe can be mapped on top of X without much trouble.
But, if you are working trying to get the most out a low end graphics
machine such as typical 486 class machines the transport overhead
involved in using most X systems will eat you alive.
Just rambling...
I once was part of a project that implemented a subset of Xlib
functionality (including input) as a library that talked directly to
the frame buffer, keyboard, and mouse (we had our reasons.) It worked
out quite nicely and had surprisingly good performance.
The main problem with X is transport overhead. Several companies now
have various flavors of "Direct X" that avoid the transport overhead.
"X" has grown enough that saying "X" no longer means just the stuff
you get for free from the X consortium.
Bob P.
--
Bob Pendleton | As an engineer I hate to hear:
bobp@hal.com | 1) You've earned an "I told you so."
Speaking only for myself. | 2) Our customers don't do that.
<<< Odin, after the well of Mimir. >>>
Technical Details
Message-ID:
<1ikvv5INNh3h@shelley.u.washington.edu>Path:
arc1-ca.pugleaf.net!sparky!uunet!zaphod.mps.ohio-state.edu!rpi!usenet.coe.montana.edu!news.u.washington.edu!stein.u.washington.edu!hlabReferences:
<1ib69aINNr9b@shelley.u.washington.edu>Show Full Headers
Complete Article Headers
Path: arc1-ca.pugleaf.net!sparky!uunet!zaphod.mps.ohio-state.edu!rpi!usenet.coe.montana.edu!news.u.washington.edu!stein.u.washington.edu!hlab From: bobp@hal.com (Bob Pendleton) Newsgroups: sci.virtual-worlds Subject: Re: TECH: Arbitrary input devices Date: Wed, 6 Jan 1993 22:44:37 GMT Organization: HaL Computer Systems, Inc. Lines: 44 Approved: cyberoid@milton.u.washington.edu Message-ID: <1ikvv5INNh3h@shelley.u.washington.edu> References: <1ib69aINNr9b@shelley.u.washington.edu> NNTP-Posting-Host: stein.u.washington.edu Originator: hlab@stein.u.washington.edu