Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!uunet!husc6!bloom-beacon!mgm.mit.edu!wolfgang
From: wolfgang@mgm.mit.edu (Wolfgang Rupprecht)
Newsgroups: comp.emacs
Subject: Re: query-replace anomaly
Message-ID: <1766@bloom-beacon.MIT.EDU>
Date: Mon, 9-Nov-87 13:34:03 EST
Article-I.D.: bloom-be.1766
Posted: Mon Nov  9 13:34:03 1987
Date-Received: Wed, 11-Nov-87 06:14:43 EST
References: <182@yetti.UUCP> <8711091504.AA25772@ucbvax.Berkeley.EDU>
Sender: daemon@bloom-beacon.MIT.EDU
Reply-To: wolfgang@mgm.mit.edu (Wolfgang Rupprecht)
Organization: Independent Software Consultant
Lines: 14

In article <8711091504.AA25772@ucbvax.Berkeley.EDU> jr@ALEXANDER.BBN.COM writes:
>No fix should be necessary ...

No there is indeed a compiler bug on some machines. (I know that at
one time Sun-3's were buggy, while Vax's were ok.) The bug causes ALL
replacements to be up-shifted if case-replace and case-fold-search are
t. I seem to recall that it was due bad code being generated for
'register char' variables on the Sun. I don't have a Sun any more, so
I can't say if it is still a problem. If Sun could provide me with a
test machine ... 	;-)
--
Wolfgang Rupprecht	UUCP: mit-eddie!mgm.mit.edu!wolfgang
				(or) mirror!mit-mgm!wolfgang
			ARPA: wolfgang@mgm.mit.edu (IP addr 18.82.0.114)