home *** CD-ROM | disk | FTP | other *** search
- Path: stdc.demon.co.uk!clive
- From: clive@stdc.demon.co.uk (Clive D.W. Feather)
- Newsgroups: comp.std.c
- Subject: Re: size_t
- Date: Fri, 2 Feb 1996 06:20:26 GMT
- Organization: Demon Internet Limited (personal account)
- Message-ID: <DM4xM4.8H6@stdc.demon.co.uk>
- References: <4eo3sc$l14@info1.sdrc.com> <4es27i$hd5@alterdial.UU.NET>
- Reply-To: clive@demon.net
- X-NNTP-Posting-Host: stdc.demon.co.uk
-
- In article <4es27i$hd5@alterdial.UU.NET>, <rex@aussie.com> wrote:
- > larry.jones@sdrc.com (Larry Jones) writes:
- >> The
- >> response to another Defect Report noted that the list of integer types
- >> is exhaustive and may not be extended,
-
- > Larry, I recall a discussion in committee re mapping pointers to some
- > integral type. Somewhere we require that I believe. IBM's AS400 came up as it
- > has (I think) 128-bit addresses but no integer type that big. I thought we
- > allowed magic integer types for some things but I may be misremembering.
-
- You're misremembering (it was *my* Defect Report, in Kona). The mapping
- from pointers to some integral type doesn't have to exist if no integral
- type is "big enough". We also said that "big enough" is an implementation
- determined concept, and is *not* a matter of comparing sizeofs. For
- example, pointers might be 32 bits, but only use 20 of them, so a 24-bit
- short might be "big enough".
-
- --
- Clive D.W. Feather | If you lie to the compiler,
- cdwf@cityscape.co.uk (work, preferred) | it will get its revenge.
- clive@stdc.demon.co.uk (home) | - Henry Spencer
-