Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/18/84; site sftig.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!mhuxj!mhuxm!sftig!pg
From: pg@sftig.UUCP (P.Glick)
Newsgroups: net.unix-wizards
Subject: smart compilers
Message-ID: <479@sftig.UUCP>
Date: Thu, 6-Dec-84 09:44:47 EST
Article-I.D.: sftig.479
Posted: Thu Dec  6 09:44:47 1984
Date-Received: Fri, 7-Dec-84 02:44:18 EST
References: <783@enea.UUCP>
Organization: AT&T Bell Laboratories, Summit, NJ
Lines: 9

I don't understand why people would consider a compiler "smart"
if it were to replace the two nested empty 'for' loops with assignments
of the loop variables.  I prefer a compiler to do the expected -- what
I told it to do.  These kinds of 'optimizations' can make programming a pain
in the neck.  When it comes to programming, I prefer to be the one that makes
the mistakes and not find out that a compiler cannot generate the code I
want it to generate.
				Paul Glick
				btlunix!pg