But does it help if you're already on a NClass and looking to define 'where
your going to be' ten years from now?  Won't the NClass be history well
before then? It may lower costs short-term, but the only real solution is
emulation on the systems that will be new ten years from now. I think it
would be nice to have too, but does it help the goal?

my 2¢ worth.
KARSTEN HOLLAND
MIS Site MGR
[log in to unmask]


-----Original Message-----
From: Sletten Kenneth W KPWA [mailto:[log in to unmask]]
Sent: Thursday, August 14, 2003 1:39 PM
To: [log in to unmask]
Subject: Re: Stock Piles


Jon after Zelik's message:
=====================================================

This does raise a question, at least in theory worth exploring.
If HP were to reverse their position on HP9000 to HP3000
conversions - for A and N class used boxes ONLY, what would be
the OpenMPE communities reaction?

How to you think it would impact your perceived need for an
emulator?  Note that even with the conversion it would still
only give us a finite amount of native environments.  Would it
help your business case for staying on the native environment
(a.k.a., HP3000) until an emulated environment is created?
=====================================================

(1)  I would expect(and hope) that the OpemMPE community would
perceive and receive any such move by HP in a positive light.
I hadn't really thought this thru until recent discussions,
but if you try and look ahead more than just a few years,
seems to me that if at all possible the newest HP e3000 A-and-N
Class boxes are really the ones sites would want to focus on if
they are expecting that they are going to be running MPE in
production for a long time yet:  The A-and-N Class are a number
of years ahead of any other 3000 family WRT to interfaces /
peripherals they support;  and just for that reason alone are
much more "supportable" in the out years.  Certainly current A-
and-N class boxes will have a "sunset" too, but I would think
much later than 9X9, etc.

(2)  WRT emulator impact:  Widespread availability of 9000-to-
3000 conversion "kits" for A-and-N would I expect almost by
definition reduce the demand for an emulator;  certainly in
the short run (however "short run" is defined).

(3)  WRT helping "business case" for staying on 3000:  It would
not help our particular site at all (for unconnected reasons),
but I would certainly expect it to help some other sites;
maybe a fairly significant number.


My SUMMARY take:  On balance, getting HP to agree to make
"9000-to-3000" conversion kits for A-and-N Class boxes
available would be a good thing...  Heck, for that matter
allowing K-Class 9000 to 9X9 3000 conversion wouldn't hurt
either:  9X9s have been around for quite a few years, but I
bet a lot of them could still be supported for quite a few
more (Keyport ran some ancient HP 1000 test set computers for
over 25 years (there were upgrades along the way, but the same
core machines were still there) ).

Ken