home *** CD-ROM | disk | FTP | other *** search
- Xref: sparky comp.edu:1573 comp.lang.fortran:3472 comp.lang.misc:2969 comp.arch:9281 sci.math:11121
- Path: sparky!uunet!olivea!spool.mu.edu!wupost!dbsun!meyer
- From: meyer@dbsun.uucp (Don Meyer)
- Newsgroups: comp.edu,comp.lang.fortran,comp.lang.misc,comp.arch,sci.math
- Subject: Re: Scientists as Programmers (was Re: Small Language Wanted)
- Message-ID: <1992Sep8.151053.3795@dbsun.uucp>
- Date: 8 Sep 92 15:10:53 GMT
- References: <Bu08uF.HBC@mentor.cc.purdue.edu> <1992Sep4.114720.25228@dbsun.uucp> <Bu2DnM.J4E@mentor.cc.purdue.edu>
- Distribution: na
- Organization: BioMerieux-Vitek, St. Louis Mo.
- Lines: 21
-
- In article <Bu2DnM.J4E@mentor.cc.purdue.edu> hrubin@pop.stat.purdue.edu (Herman Rubin) writes:
- >In article <1992Sep4.114720.25228@dbsun.uucp> meyer@dbsun.uucp (Don Meyer) writes:
- >>If you don't have floating point hardware, or if the precision need
- >>surpasses that of your floating point hardware you will almost
- >>certainly have a subroutine call made by the compiler anyway.
- >>This is an optimization/compiler-quality issue as opposed to
- >>a pure "language" issue IMHO.
- >
- >This is not necessarily the case. At least on some machines, one can
- >handle twice that precision without calling a subroutine for floating
- >point operations. For integer even triple and quadruple precision
- >often should be programmed without a subroutine call.
-
- That's what I said, isn't it? It's heavily dependant on the hardware,
- as well as the speed versus code-size tradeoff desired.
-
- Don
- --
- -------------------------------------------------------------------------
- meyer%dbsun.uucp@wupost.wustl.edu ||| I am at two with nature.
- Opinions expressed are personal. ||| -- Woody Allen
-