Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/18/84; site pokey.UUCP
Path: utzoo!watmath!clyde!cbosgd!cbdkc1!desoto!packard!hoxna!houxm!mhuxj!mhuxr!ulysses!allegra!mit-eddie!genrad!panda!talcott!harvard!seismo!uwvax!gumby!pokey!tamayo
From: tamayo@pokey.UUCP
Newsgroups: net.lang,net.lang.mod2
Subject: Re: nil pointer dereferencing
Message-ID: <227@pokey.UUCP>
Date: Tue, 26-Feb-85 16:22:05 EST
Article-I.D.: pokey.227
Posted: Tue Feb 26 16:22:05 1985
Date-Received: Fri, 1-Mar-85 06:21:59 EST
References: <286@gumby.UUCP>
Distribution: net
Organization: U of Wisconsin CS Dept
Lines: 25
Xref: watmath net.lang:1468 net.lang.mod2:180

> 
>    Since I don't remember what newsgroup this discussion was in, I'll
> try these two . . .
> 
>    The Logitech Modula-2/86 compiler for the 8086, as of version 1.1,
> has an address constant called NIL which produces the value FFFF:FFFF
> (segment:offset).  Since this is not a very valid address on the 
> processor (it wraps around into lower memory), no addressability is
> lost by this convention.
> 
>    Actually, this was in v1.0.  What is new in 1.1 is that, by default,
> the compiler generates code to check for a NIL dereferencing operation
> before it takes place.  This checking may be turned off by a compile-

> 
> 
> -- 
>       Dan Frank
> 
> 	  Q: What's the difference between an Apple MacIntosh
> 	     and an Etch-A-Sketch?
> 
> 	  A: You don't have to shake the Mac to clear the screen.

*** REPLACE THIS LINE WITH YOUR MESSAGE ***