home *** CD-ROM | disk | FTP | other *** search
- Path: airdmhor.gen.nz!not-for-mail
- From: gumboot@airdmhor.gen.nz (Simon Hosie)
- Newsgroups: comp.lang.c
- Subject: Re: Defining accuracy
- Date: 26 Mar 1996 06:45:35 +1200
- Organization: Airdmhor : a couple of BBS's, a bunch of people, and a cat.
- Message-ID: <4j6pkf$e68@airdmhor.gen.nz>
- References: <DoKBM0.8L3.B.midge@bath.ac.uk> <827684758snz@genesis.demon.co.uk>
- NNTP-Posting-Host: airdmhor.gen.nz
- X-Newsreader: TIN [version 1.2 PL2]
-
- Lawrence Kirby:
- > In article <DoKBM0.8L3.B.midge@bath.ac.uk> py3heh@bath.ac.uk "H E Harmer" writes:
-
- > >Is there a way of defining the precision of the 'double' and 'long double'
- > >variables when compiling with an ANSI C compiler? I am looking to
- > >hopefully doubling it or better.
-
- > Generally no - the compiler defines the precision of the various floating
- > point types. It typically corresponds to what the hardare supports.
-
- Couldn't you check the values in limits.h and typedef the type that fits,
- or give an error or take the best fit if you can't find anything?
-