Path: utzoo!utgpu!jarvis.csri.toronto.edu!clyde.concordia.ca!uunet!zds-ux!gerry
From: gerry@zds-ux.UUCP (Gerry Gleason)
Newsgroups: comp.arch
Subject: Re: Byte ordering
Message-ID: <148@zds-ux.UUCP>
Date: 2 Feb 90 19:09:22 GMT
References: <9656@spool.cs.wisc.edu>
Reply-To: gerry@zds-ux.UUCP (Gerry Gleason)
Organization: Zenith Data Systems
Lines: 21

In article <9656@spool.cs.wisc.edu> rbi@beaufort.cs.wisc.edu (Bruce Irvin) writes:
>  Could someone send me (or post) a brief summary of the 
>major issues in choosing Big or Little Endian byte ordering 
>for an architecture?

Pretty much everyone agrees that it is a wash with respect to performance,
functionality, and the other important issues, except for a few zealots.
The main issue is which of a few tricks that can be pulled:  on a big-
endian machine, you can compare alligned (and padded) strings word at a
time instead of byte by byte, or on a little endian that byte, half-word
or word accesses through the same pointer will get the same value assuming
you haven't truncated any important bits.  Of course, both of these are
tricks, and the second could be considered a disadvantage because it makes
it more likely that incorect code might work much of the time.

Btw, has anyone else seen the extended rhyme on this topic.  It's a takeoff
on Swift's satirical piece about which end an egg should be opened from
(it's in Guliver's Travels if I'm not mistaken).  If someone has it, could
you email or post it.

Gerry Gleason