Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!watmath!clyde!rutgers!sri-spam!sri-unix!hplabs!tektronix!uw-beaver!ubc-vision!ubc-cs!majka
From: majka@ubc-cs.UUCP
Newsgroups: net.graphics
Subject: Re: Y positive up or down?
Message-ID: <451@ubc-cs.UUCP>
Date: Fri, 24-Oct-86 13:33:20 EST
Article-I.D.: ubc-cs.451
Posted: Fri Oct 24 13:33:20 1986
Date-Received: Mon, 27-Oct-86 00:34:47 EST
References: <469@vaxb.calgary.UUCP>
Reply-To: majka@ubc-cs.UUCP (Marc Majka)
Distribution: net
Organization: UBC Department of Computer Science, Vancouver, B.C., Canada
Lines: 23
Keywords: cartesian coordinates, pixel addressing

I went through this debate with several users as I was implementing a set of
graphics utilities such as you mention.  I was gung-ho on Row-Column coordinate
systems (origin upper left, first coordinate increasing down).  Some users
concurred but others argued for cartesian X-Y spaces.  In the end I decided on
the philosophy of flexibility:

  Let The Users Choose The Coordinate System That They Are Happy Working With.

This solution caused me to do a bit more work in writing the coordinate 
transformation routines, but it was well worth the small amount of effort.
The users just begins with a specification of the coordinate space they want.
In fact, they can change to a new coordinate system anytime.  This can be very
handy if you are working with a mixed display.  For instance where you want to
do some x-y plotting, and add a lot of text annotation.

There are always some cases where you will want to work in different coordinate
systems.  You may want the origin anywhere on the display.  You may want the
coordinates to run in either direction.  It is not hard to let the utility do a
bit more work for you.  After all, computers should make life *easier* for
people!

---
Marc Majka  -  UBC Laboratory for Computational Vision