Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!csd4.milw.wisc.edu!lll-winken!xanth!mcnc!ecsvax!dukeac!wolves!ggw From: ggw@wolves.UUCP (Gregory G. Woodbury) Newsgroups: comp.lang.c Subject: Re: long double Summary: MAC short double Message-ID: Date: 1 May 89 01:16:52 GMT References: <12787@haddock.ima.isc.com> <12661@lanl.gov> <1989Apr26.171042.4029@utzoo.uucp> Organization: Red Wolfe Software Lines: 19 In <1989Apr26.171042.4029@utzoo.uucp> henry@utzoo.uucp (Henry Spencer) wrote: > In article <12661@lanl.gov> jlg@lanl.gov (Jim Giles) writes: > >By the way, what _is_ "long double"? I've never seen a C compiler which > >has such a thing. The proposed ANSI standard mentions "long double", > >but doesn't require it to be more precise than double! ... > > The proposed standard also doesn't require float to be less precise than > double. The reason is the same: the exact set of easily-supported sizes > is machine-specific. Providing float lets the user get at a smaller type, > IF there is one. Providing long double lets the user get at a larger > type, IF there is one. Some machines have all three. Many don't. For wierdness: Think C (LightSpeed C) has these floating point types: float 4 bytes double 10 bytes short double 8 bytes The manual notes that the short double is provided because the 8 byte values are the slowest to compute with in the SANE 68881 environment (macintosh).