Xref: utzoo comp.arch:7750 comp.misc:4590 comp.lang.misc:2435 comp.protocols.misc:434
Path: utzoo!utgpu!watmath!clyde!att!rutgers!mailrus!cornell!uw-beaver!microsoft!w-colinp
From: w-colinp@microsoft.UUCP (Colin Plumb)
Newsgroups: comp.arch,comp.misc,comp.lang.misc,comp.protocols.misc
Subject: Re: "big endian" and "little endian" - first usage for computer
Message-ID: <182@microsoft.UUCP>
Date: 4 Jan 89 22:17:08 GMT
References: <170@microsoft.UUCP> <4008@hubcap.UUCP>
Reply-To: w-colinp@microsoft.UUCP (Colin Plumb)
Organization: Microsoft Corp., Redmond WA
Lines: 30

In article <4008@hubcap.UUCP> hutch@delft writes:
>Actually, where most little-endian machines screw up is storing the
>bits in the byte in the wrong order.  It is good to hear that somebody got it
>right and stored a one as 100000...0000 rather than 00000001000...000.
>(That is what you meant wasn't it?).  Note that this implies one
>multiplies by 2 by using a RIGHT shift (else there is an inconsistancy
>in the little-endian view in the registers)!  The Inmos sounds interesting.

Sorry, no.  Little endian means that if two addressed objects (on the
Trasnputer, the smallest object that can be addressed is a byte) are
part of the same number, the object (byte) with the lower address is
less significant.

Note two things:
-> There is no ordering implied on bits within bytes; a byte is an atomic
   object, and you can't say which bit of it comes "first."  (Of course,
   in serial communications, the other site of the Holy War, this is
   significant.)
-> Both big- and little-endian types agree that more significant bits should
   be to the left, conceptually (the Arabic heritage, remember?); they
   *don't* agree on whether addresses increase left-to-right (big-endian) or
   right-to-left (little-endian).  See On Holy Wars and a Plea For Peace for
   more diagrams.  A one is stored as 00000000 00000000 00000000 00000001,
   with the bytes' addresses being     base+3   base+2   base+1   base+0.

Thus, it is impossible for a byte-addressed machine to store the bits in
a byte in the wrong order, unless it has bitfield instructions or some
such bit-addressing kludge.
-- 
	-Colin (uunet!microsof!w-colinp)