Path: utzoo!utgpu!news-server.csri.toronto.edu!bonnie.concordia.ca!thunder.mcrcim.mcgill.edu!mouse
From: mouse@thunder.mcrcim.mcgill.edu (der Mouse)
Newsgroups: comp.windows.x
Subject: Re: (none)
Message-ID: <1991Jun30.215955.6388@thunder.mcrcim.mcgill.edu>
Date: 30 Jun 91 21:59:55 GMT
References: <9106282259.AA24463@rhine.awdpa.ibm.com>
Organization: McGill Research Centre for Intelligent Machines
Lines: 29

In article <9106282259.AA24463@rhine.awdpa.ibm.com>, wsh@ibmpa.awdpa.ibm.COM (Bill Hughey) writes:

> I will be doing a port of X that will take advantage of a graphics
> accelerator device.

> 1. From the code it seems that widelines need floating point up to
>    span generation thus ruling out using an accelerator that only
>    uses integer input parameters.  True/False ?

False.  The current code may use floating point, but there is no need
for anything but integer operations.  (All coordinates can be expressed
as rational numbers - ie, as one integer divided by another - so they
can be dealt with using purely integer arithmetic.  The only thing you
need to watch out for is overflow.)

> 2. Are there any public domain codes for decomposing general polygons
>    into triangles or quadrilaterals for accelerators that can't do
>    polygons with more than 4 corners?

I don't know the answer to this, but there are certainly computational
geometry texts that will contain algorithms for triangulating polygons.
Implementing one of those may well be easier than doing it all
yourself.

					der Mouse

			old: mcgill-vision!mouse
			new: mouse@larry.mcrcim.mcgill.edu