Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!tut.cis.ohio-state.edu!quanta.eng.ohio-state.edu!sleepy.eng.ohio-state.edu!rob
From: rob@sleepy.eng.ohio-state.edu (Rob Carriere)
Newsgroups: comp.lang.c
Subject: Re: Character constants
Keywords: NULL
Message-ID: <1586@quanta.eng.ohio-state.edu>
Date: 3 Mar 89 01:40:45 GMT
References: <13068@steinmetz.ge.com> <102@servio.UUCP> <10138@socslgw.csl.sony.JUNET> <1783@dlvax2.datlog.co.uk> <217@skye.ed.ac.uk> <7447@pyr.gatech.EDU>
Sender: news@quanta.eng.ohio-state.edu
Reply-To: rob@sleepy.eng.ohio-state.edu (Rob Carriere)
Organization: Ohio State Univ, College of Engineering
Lines: 22

In article <7447@pyr.gatech.EDU> curci@stat.fsu.edu (Ray Curci (scri)) writes:
>Why is therer so much discussion about  character discussions?
>
>A simple printf("%d", sizeof('x') ); reveals that a character constant
>(at least both on sun3 and ultrix 11/780) are the same as an int (4 chars).
>Isn't it a lot easier to execute a simple test program than to try to
>look up this type of information?

Yes, it is (usually) much easier.  Yes, you should do it, *to test
your compiler*.  No, it does not answer the question.  A compiler does
not a language definition make.

Secondly, your test here is insufficient.  For example, the computer
that is currently bearing with me chattering all over the place will,
on request, also happily provide the information that

sizeof(long) == sizeof(float) 

The assumption that this would mean that the types long and float are
the same seems rather unsafe.

SR