Path: utzoo!attcan!uunet!lll-winken!lll-tis!helios.ee.lbl.gov!pasteur!agate!bionet!apple!bloom-beacon!gatech!uflorida!haven!vrdxhq!bms-at!stuart
From: stuart@bms-at.UUCP (Stuart Gathman)
Newsgroups: comp.lang.c
Subject: Re: Something IBM did right (RT division of negatives).
Summary: An actual situation
Message-ID: <124@bms-at.UUCP>
Date: 8 Nov 88 16:44:03 GMT
References: <3105@hubcap.UUCP> <34112@XAIT.XEROX.COM> <1700@dataio.Data-IO.COM> <11529@bellcore.bellcore.com>
Organization: Business Management Systems, Inc., Fairfax, VA
Lines: 13

In article <11529@bellcore.bellcore.com>, sjs@jcricket.ctt.bellcore.com (Stan Switzer) writes:

> Anyone wishing to take up the other side of the argument must find an
> example of an actual situation where having -1/2 yield -1 is useful.

How about rounding by adding 1/2 and truncating (toward the smallest number,
not zero)?

Trunctating (-1)/2 to -1 is exactly the approach that does *not* require
checking the sign when using arithmetic shifts.  If you want logical
shifts, use unsigned operands.
-- 
Stuart D. Gathman	<stuart@bms-at.uucp>
			<..!{vrdxhq|daitc}!bms-at!stuart>