🚀 go-pugleaf

In article <RYAN.91Jul17225230@ra.cs.umb.edu> ryan@ra.cs.umb.edu
(Daniel R. Guilderson) writes:

>In article <1738@opus.NMSU.Edu> sitze@nmsu.edu (Richard Sitze) writes:
>
>   My only complaint is against those who feel that the idea is to
>   OBVIOUS.  Yes, it is simple and a beautiful solution to the problem.
>   But how many figured it out for themselves??  Like many of the BEST
>   ideas, it's only to obvious once you know it.
>
>Your assignment for the day, should you choose to accept it Mr.
>Phelps, is to transfer exactly one 32-bit word between a VAX and a
>68000 computer.  The word must represent the same integer on both
>architectures.  How do you do it without byte-swapping?

You are looking at this from a narrow point of view.

By byte-swapping (as done by Sun NFS (and thus I assume the patent
that Sun has supposedly infringed on declares a similar approach))
I would transfer a 32-bit word from a VAX to an Intel 386 by
converting VAX ordering to Motorola (NFS standard matches the Motorola
68000 byte ordering I recall), send it over the network to the Intel 386,
and convert to local Intel byte ordering - a round about way one must
admit, and certainly not the obvious solution to your original
question (though 'prehaps' obvious when one looks at the larger
picture).  Even worse, if I follow the NFS standard, I send from VAX A
to VAX B by converting to Motorola 68000 as a middle step....

(Am I right in recalling that the VAX 32 bit word is slightly
different than the Intel 32 bit word??? - it's been a long time, but
the argument stands in either case...).

	<ras>

--
                            +--------------------------
                            | Richard A. Sitze
                            | sitze@nmsu.edu,    phone: (505) 646-6228
                              SH 163