Re: A3000 -> transfers

Graham Barlow (gkb1@tower.york.ac.uk)
Tue, 26 Oct 1993 16:55:02 +0000

To: marat@CC.UManitoba.CA
From: gkb1@tower.york.ac.uk (Graham Barlow)
Subject: Re: A3000 -> transfers
Cc:
Bcc:
X-Attachments:

>Hi!
>
>A problem arises when transfering data from a word addressed
>computer like the A3000 to a byte addressed computer (eg a
>UNIX box). That is, when the 24 bit word is split into bytes, which
>byte is sent first?

This problem has been eddressed by several people before, don't rush to
write your own code just yet as this has been done (many times, I'm sure)

>Does anyone out there know if there are byte order differences
>in the resulting UNIX file when an A3000 file is transfered to
>a UNIX computer by different methods?

Yes! The files resulting from a transfer by Kermit are different to those
by Bruknet.

>ie. Does a Bruknet transfer produce a different byte order
>than, say, a Kermit transfer? Do versions of bruknet
>running on big endian machines like a Sun or SGI result in
>the same byte order as bruknet on the little endian X32?
>
>
I once wrote a VAX Pascal program for Kermitted data and I know of at least
one C source. If you would like I will dig out some bits ad pieces. By the
way, what exactly do you need to do?

Regards, Graham
+-----------------------===============================================+
| Dr. Graham Barlow | tower or VAX mail: gkb1 |
| NMR Service Manager | JANET mail: gkb1@UK.AC.YORK |
| Dept. of Chemistry | Internet mail: gkb1@york.ac.uk |
| University of York | Telephone: +44 (0)904 432506 DDI |
| Heslington, YORK, UK | FAX: +44 (0)904 432516 |
+-----------------------+==============================================+