Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!watmath!clyde!caip!rutgers!sri-spam!sri-unix!hplabs!hpcea!hpfcdc!hpcnoe!jason
From: jason@hpcnoe.UUCP (Jason Zions)
Newsgroups: net.lang.c
Subject: Re: Signed char - What Foolishness Is This!
Message-ID: <790009@hpcnoe.UUCP>
Date: Sat, 18-Oct-86 15:08:03 EDT
Article-I.D.: hpcnoe.790009
Posted: Sat Oct 18 15:08:03 1986
Date-Received: Tue, 21-Oct-86 22:03:10 EDT
References: <8719@duke.duke.UUCP>
Organization: HP Colorado Networks Division
Lines: 40

jwg@duke.UUCP (Jeffrey William Gillette) /  6:26 am  Oct 16, 1986 /

> 1)      Do other C compilers make 'char' a signed quantity by default?

Oh, let's see. The PDP-11 and VAX chars are signed by default. There are
doubtless many others; I believe 68000 defaults to signed as well.

> 2)      What possible justification is there for this default?

You'll notice in the above list that I mentioned CPU types, rather than
particular compilers. That's because of the answer to your question 2. A
compiler may choose to make chars signed or unsigned depending upon which
is more efficient. chars are indeed arithmetic objects; if they weren't,
you couldn't freely intermix them with integers.

The problem you're running into is an age-old one; the automatic conversion
of a char to an int. If a char is signed, the conversion is a sign-extending
one, resulting in a potentially negative int. If the char is unsigned, the
conversion is (essentially) a bit-field copy, and the resulting int is
always positive.

K&R, chapter 2, page 40:

	The language does not specify whether variables of type *char*
	are signed or unsigned quantities. When a *char* is converted
	to an *int*, can it ever produce a *negative* integer? Unfortunately,
	this varies from machine to machine, reflecting differences in
	architecture.

Have fun; this one bites everyone at least once... I've lost about a pound
of flesh from a sensitive portion of my anatomy from being bitten so often.
--
This is not an official statement of Hewlett-Packard Corp., and does not 
necessarily reflect the views of HP. It is provided completely without warranty
of any kind. Lawyers take 3d10 damage and roll a saving throw vs. ego attack.

Jason Zions				Hewlett-Packard
Colorado Networks Division		3404 E. Harmony Road
Mail Stop 102				Ft. Collins, CO  80525
	{ihnp4,seismo,hplabs,gatech}!hpfcdc!hpcnoe!jason