Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1 6/24/83; site mit-hermes.ARPA
Path: utzoo!watmath!clyde!cbosgd!ihnp4!mhuxn!mhuxr!mhuxt!houxm!whuxl!whuxlm!harpo!decvax!genrad!mit-eddie!mit-hermes!ads
From: ads@mit-hermes.ARPA (Sathya D.Narayanan)
Newsgroups: net.math
Subject: Fraction math algs.
Message-ID: <2378@mit-hermes.ARPA>
Date: Sat, 27-Apr-85 15:14:26 EDT
Article-I.D.: mit-herm.2378
Posted: Sat Apr 27 15:14:26 1985
Date-Received: Sun, 28-Apr-85 23:48:01 EDT
Organization: The MIT AI Lab, Cambridge, MA
Lines: 13


	The Art of Computer Programming,Vol 2 by D.E.Knuth on pg 315
lists the references.Read the one's by Matula and Kornerup this should
give all the details you need.
	Though don't expect it to be faster than floating point
especially if you are on a machine which has floating point
hardware.If the machine has floating point hardware then you should
maybe just use it if you aren't too concerned about exact results.The
rational stuff could give you exact results since you never have to
approximate,provided you represent both numerator and denominator as
large enough integers.

-ads