Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!utgpu!water!watmath!clyde!rutgers!rochester!PT.CS.CMU.EDU!F.GP.CS.CMU.EDU!rpd
From: rpd@F.GP.CS.CMU.EDU.UUCP
Newsgroups: comp.lang.c
Subject: Re: Question on structures
Message-ID: <295@PT.CS.CMU.EDU>
Date: Tue, 3-Nov-87 16:47:48 EST
Article-I.D.: PT.295
Posted: Tue Nov  3 16:47:48 1987
Date-Received: Sat, 7-Nov-87 08:13:01 EST
References: <1025@phoenix.Princeton.EDU> <2986@mit-vax.LCS.MIT.EDU> <5673@weitek.UUCP> <3048@psuvax1.psu.edu>
Sender: netnews@PT.CS.CMU.EDU
Organization: Carnegie-Mellon University, CS/RI
Lines: 25

The topic under discussion is the Pascal WITH statement and C
equivalents.  Dan Rober points out that, in C, it is easy to declare
a local pointer variable and initialize it with the address of the
structure to be manipulated.  He claims that this approach avoids
ambiguities present in the WITH statement.  Scott Schwartz isn't
convinced and feels that "Anyone framiliar [sic] with nested block
structured languages should have no problem with this."

I have to agree with Dan's position.  In my experience, every WITH
statement is a mystifying bug waiting to happen.  Consider the following
fragment of code:
	VAR foo : integer;
	BEGIN
		WITH a.b.c DO
			foo = bar + baz
	END;
where bar & baz are fields of a.b.c.  This is code buried in some obscure
long-forgotten part of your program.  Now you decide to add a new field
to a.b.c, and call it foo.  All of a sudden that old code which you didn't
touch and certainly didn't think to look at stops working.

The upshot is that by using WITH the programmer throws away the advantages
that block structure is supposed to give him.

Rich