Path: utzoo!attcan!utgpu!jarvis.csri.toronto.edu!mailrus!ncar!ames!elroy.jpl.nasa.gov!decwrl!granite!aps From: aps@granite.dec.com (Armando P. Stettner) Newsgroups: comp.sys.mips Subject: Re: Clarification Message-ID: <516@granite.dec.com> Date: 29 Jul 89 19:42:39 GMT References: <2293@itivax.iti.org> <23924@winchester.mips.COM> Reply-To: aps@granite.UUCP (Armando P. Stettner) Distribution: comp Organization: DEC Workstation Systems Engineering Lines: 35 In article rich@sendai.ann-arbor.mi.us writes: ..... >Then I'd like to suggest that: > .... > >b) someone yell at DEC and tell them that their machine isn't a >"mips". Specifically, the DEC distributed C compiler defines "mips", >"MIPSEL", and "ultrix", which is your best clue to the fact that this >is a decstation, but machine(1) reports that the decstation is a >"mips". >-- >rich. Our view is that machine(1) and the "mips" defination in the C compiler are indications of the architecture; not the specific model in a family. For example, you don't get an indication that the machine is a 780 or 750 or 6200; just that it is a VAX. If you have to know which machine it is, you are probably writing kernel code and there are other methods used there to differentiate models and characteristics. As for the "MIPSEL" and "ultrix", the former is an indication from the compiler as to which endianism the target machine is and the later indicates the target operating system. Both of these are defaults in the compiler and MIPSEL is also used by other (all?) MIPS C (other?) compilers. This complaint/comment should have gone to net.whatever for Digital systems and/or Ultrix. Sorry to take up the news message, John. armando. decwrl!aps aps@decwrl.dec.com