Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!wuarchive!zaphod.mps.ohio-state.edu!pacific.mps.ohio-state.edu!linac!att!pacbell.com!ames!haven!decuac!decatl!wreck.alf.dec.com!hoyt From: hoyt@wreck.alf.dec.com (Kurt Hoyt) Newsgroups: comp.unix.ultrix Subject: Re: Sun (etc) ieee -> DS ieee Message-ID: <1990Nov27.092900@wreck.alf.dec.com> Date: 27 Nov 90 14:29:00 GMT References: <1395@mpirbn.mpifr-bonn.mpg.de> <1990Nov24.201244.7225@ircam.fr> Sender: news@decatl.dec.com (USENET News System) Reply-To: hoyt@wreck.alf.dec.com (Kurt Hoyt) Organization: Digital Equipment Corporation Lines: 25 Nntp-Posting-Host: wreck.alf.dec.com In article <1395@mpirbn.mpifr-bonn.mpg.de>, p554mve@mpirbn.mpifr-bonn.mpg.de (Michael van Elst) writes: |> In article <1990Nov24.201244.7225@ircam.fr> mf@ircam.fr (Michel Fingerhut) writes: |> >Use xdr - library routines for external data representation. |> |> Except for xdr_float/xdr_double is broken in Ultrix3.1 This problem has been fixed in ULTRIX 4.0. The xdr_float.c file has ifdefs to tell it whether you are using VAX or IEEE floating point. In xdr_float and xdr_double, the floating point format was determined by this construct: #ifdef mc68000 ...IEEE code... #else ...VAX code... #endif If you weren't a Motorola 68000, the routine thought you were a VAX. Since the MIPS chip is not a 68000, xdr_float incorrectly considered it a VAX. -- Kurt in Atlanta | Atlanta hoyt@decatl.alf.dec.com | Site of the 1996 Olympics hoyt@wreck.alf.dec.com |