Stigers, Greg [And] <[log in to unmask]> wrote:
>
> How "proprietary" is MPE? It both is and isn't. Our Oracle DBAs frequently
> just use familiar UNIX / POSIX commands from the CI, and will even work
> entirely in the shell. But MPE * is * proprietary. Does that doom it to
> eventual extinction? I think that there is at least one list member who
> understand extinction rather well, who would disagree. Proprietariness is
> not so much hurting the UNIX / Linux variants, nor Microsoft, nor IBM's
> OS/390 or OS/400 (which doesn't even have Apache or SAMBA at this time!).
> Why should it hurt MPE? It's a state of mind, addressed by marketing! It
is
> very much HP's choice if they want to be like Unisys, Wang, Data General,
> DEC, Prime, or like NT / W2K and the AS/400.

IBM and Microsoft will survive, other than those, there is only Unix.
(And IBM is consolidating to OS/390, OS/400, and Unix, discouraging new
users on VM/CMS, VSE, OS/2 and anything else that may still linger.)

The same consolidation is happening in the Unix market.  Only 4 proprietary
Unix variants are still viable--HP, Sun, IBM, DEC/Compaq.  And I think
DEC/Compaq's proprietary Unix may not have long left.  Other computer makers
are either adopting one of these Unixes or another "standard" unix, going to
a niche strategy, or disappearing.  SCO has collapsed.

Most computer makers are choosing Linux as one supported OS, and usually
choose Red Hat Linux as the particular variant to support, because it is the
most popular.  HP, Compaq, and IBM are all supporting Linux (Sun does sell
Linux for its machines, but doesn't really support it...yet). This trend
will continue.

Linux is in that dynamic state where market share can change very fast.  As
a result, hardware and software vendors are making the commitment to get in
early--Oracle and other DBMSs are on Linux, IBM is porting its Logical
Volume Manager, SGI is porting its XFS--both of which will make Linux a much
more powerful system.  It's going to be a pain to administer for quite a
while, but clearly Linux is the direction the industry is moving.


>
> Ken wrote:
> > Occassionally there will be a new technology that allows newcomers to
> enter
> > the market: minicomputers, microprocessors, local area networks, the
> > internet.  Thus we had DEC, Microsoft, Apple, Compaq, etc.
>
> The challenge is attracting new users and new customers. Someday,
newcomers
> may outgrow their lower-end solutions. After they have clustered together
a
> bunch of Linux or NT / W2K boxes, how happy are they and their customers
> with that? Did they do a great job with a great product, or is their
> always-on infrastructure always on fire, and some part of your support
staff
> is always spending its time fighting fires? At what point do these
companies
> ask if there is a better platform, and what that might be? I think that
this
> is part of HP's strategy with Linux, in the belief that customers will
> outgrow that approach, and what a bigger box running a more slowly
changing
> OS. Can the e3000 be that box? Why or why not?

No.  There is very little reason for new customers to move to the 3000.
They would be locked into one hardware vendor and have very little choice of
software (no Oracle, e.g.).  There is no way for HP to attract software
vendors to the native 3000 interface.  HP could add on all the features to
allow all the Unix applications to run on MPE.  Then MPE would have all the
same features _and problems_ as HP-UX.  Why would a customer choose MPE over
HP-UX? Or Solaris? Or AIX?

If MPE were available as a little add-on to HP-UX, would you use it there?
No?  Then why would people using Unix applications want to run them on MPE?

If they want to move away from Unix applications, why MPE rather than
AS/400?

Sorry, but the moment has passed.  At one time, MPE was viable.  I cannot
see any way now.  Clearly, HP management has come to the same conclusion.