NetNews Offline 2/NetNews Offline Volume 2.iso/news/comp/lang/c-part2/14882

Internet Message Format | 1996-08-05 | 1.7 KB

Path: mail2news.demon.co.uk!genesis.demon.co.uk
From: Lawrence Kirby <[e-mail]>
Newsgroups: comp.lang.c
Subject: Re: Largest effective integer.
Date: Mon, 15 Apr 96 23:36:44 GMT
Organization: none
Message-ID: <[e-mail]>
References: <4ke5gu$[e-mail]> <[e-mail]> <4kqtii$[e-mail]>
Reply-To: [e-mail]
X-NNTP-Posting-Host: genesis.demon.co.uk
X-Newsreader: Demon Internet Simple News v1.27
X-Mail2News-Path: genesis.demon.co.uk

In article <4kqtii$[e-mail]>
           [e-mail] "Simon Hosie" writes:

>@#$%!?!:
>> Measly little 32 bit integers and reals. I grew up on 60 and 48 bit integers,
>> 120 bit doubles. And 65535 bit integers.
>
>> In other words, this all machine specfic. Look in <limits.h> if your
>> compiler has one.
>
>  Yes, but char short int and long have minimums defined for them, I was
>hoping that there is a minimum for the mantissa of a float or double.

The mantissa represents a value between 0 and 1. The standard does define
minimum limits on its precision (well roughly in terms of the number
of decimal digits that can distinguished):

 FLT_DIG       6
 DBL_DIG       10
 LDBL_DIG      10

These don't guarantee that integers can be represented exactly though.

>  Is long double an ANSI type, by the way?  I didn't think it was, but
>Watcom supports it even though it's the same size as a double.

long double is an ANSI defined type. E.g. a standard printf and scanf
conversion specifier for it is %Lf

--
-----------------------------------------
Lawrence Kirby | [e-mail]
Wilts, England | [e-mail]
-----------------------------------------