Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.1 6/24/83; site opus.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!harpo!seismo!hao!cires!nbires!opus!rcd
From: rcd@opus.UUCP
Newsgroups: net.unix-wizards
Subject: Why Berkeley changes in ctype?
Message-ID: <190@opus.UUCP>
Date: Mon, 5-Mar-84 22:07:38 EST
Article-I.D.: opus.190
Posted: Mon Mar  5 22:07:38 1984
Date-Received: Wed, 7-Mar-84 06:42:08 EST
Organization: NBI, Boulder
Lines: 18

(Apologies if this has been answered before.)

Does anyone know why the 4.1a and later Berkeley versions of ctype_.c (in
libc/gen) were changed so that the characters HT, NL(LF), VT, NP(FF), and
CR (011-015) are no longer "control characters"?  For example, given
	#define	HT	'\011'
the test
	if (iscntrl(HT)) . . .
will fail in the Berkeley world.

This seems to have been changed at Berkeley; I believe it happened between
4.1 and 4.1a.  It puts the Berkeley "ctype" at odds with V7, System III
and V, the manual page ctype(3), and the ASCII standard (X3.4-1977) - in
other words, the rest of the world.  It can really bite you where you sit,
since it leaves some 7-bit codes not covered by either isprint or iscntrl.
-- 

{hao,ucbvax,allegra}!nbires!rcd