HP3000-L Archives

April 2004, Week 3

HP3000-L@RAVEN.UTC.EDU

Options: Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
Mark Landin <[log in to unmask]>
Reply To:
Mark Landin <[log in to unmask]>
Date:
Fri, 16 Apr 2004 12:04:41 -0500
Content-Type:
text/plain
Parts/Attachments:
text/plain (32 lines)
 On Thu, 15 Apr 2004 15:16:45 -0500 (Central Daylight Time), Tracy
Pierce <[log in to unmask]> wrote:

>the question was "possibility", not "probability", and our friend Alfredo
>tends to answer in the most positive mindset possible.
>
>Yes it is possible.  It's even possible that HP's real reason for their
>abandonment without letting go is...
>
>MPE would probably fit atop Unix a lot better than the other way around.
>CSY probably always thought so too, but when MPE/iX happened, I'm sure it
>was a lot less risky to put iX features atop MPE.

I don't know that I would agree. We have a Poxis shell on MPE which is
a lot of what a "UNIX on MPE" would involve. However, there are just
some facilities that the MPE "kernel" requires that the UNIX kernel
does not provide. For instance, there is no facility in the HP-UX
kernel to lock a file. There are a number of user-implemented ways to
lock a file, but it's really on the honor system ... no locking is
enforced by the O/S. Shoot, you can delete open files on UNIX from
underneath people. UNIX doesn't even understand "records" ... just a
bag-of-bytes.

UNIX really has some weird idiosyncracies that date all the way to
it's original design. In many ways, it would be a foundation of sand
on which to build a robust thing like MPE. Either the kernel itself
would need some extensive reworking, or some of the features that make
MPE so good would have to be abandoned.

* To join/leave the list, search archives, change list settings, *
* etc., please visit http://raven.utc.edu/archives/hp3000-l.html *

ATOM RSS1 RSS2