Fellow listers,

[Sorry guys, I think I have gotten a bit long winded here, but you are
forewarned!]

Yesterday I started a thread by sharing with you the contents of the latest
speech by Carly.  I did not intend to reignite the firestorm that occurred.
I think the last line of my post could have been worded somewhat
differently.  I wrote "Is this the death knell from Carly for MPE or a
challenge to greater openness, increased emphasis on Posix?"  I honestly
intended it to be a challenge to see what we as users could do to help HP
make the 3000 a more open, saleable box.  There were a few that picked up on
it, but most ran with the negativity that I started.  I had not intended the
piece to be negative, but in retrospect, it certainly was.

Someone once more raised the issue of an ad in the WSJ.  I think it's a
great idea, but it doesn't sit well with everyone.  It was countered with a
proposal to spend that money either on a charity cause or to develop new
applications for the HP e3000.  I am not sure how much of an application
(much less an application suite) could be developed for the cost of an ad.
I would proffer... not much.  I am not even sure that you could port many of
the available solutions to the 3000 for that kind of money.  As to charity,
I am not sure we could all agree on a common charity, so that would best be
addressed on an individual basis at home.

I agree with Duane that HP is getting killed in the high end server market.
However, it is my perception that they are also getting killed in the
mid-range systems (where the e3000 lives), with the exception of the HP-UX
platform.  The main competition - the AS/400 - continues to be one of IBM's
best selling product lines.  Developers are still clamoring at the door to
develop for the AS/400 platform despite the fact that it is a proprietary
system also.  This does not bode well for those who are trying to make the
3000 platform a reasonable choice.

Duane commented that HP is answering the needs of the "Amazon.com"s of the
world by offering a solution that they can use.  That seems to be a
short-sighted approach as well.  First, there are relatively few companies
(in the grand scheme of things) that need and can afford a box the size of
the SuperDome.  The average business operates on a much smaller scale.
While the margins may be thinner on a box the size of a high-end HP e3000,
the collective margins on the quantity of boxes that could be sold if the
platform is positioned correctly should exceed the total margin of the
collective margins of SuperDome.

In his parting words, Duane challenged that we should "continue to write
good systems that your customers want so they will continue to buy/upgrade
to new e3k boxes as they become available".  I think that is only part of
the solution.  True, we who currently write systems for the HP e3000 box
need to write excellent (not just good) systems for the 3000.  It will
ensure that our customer base stays stable, and perhaps we can get some
referral business from the satisfied user community of our respective
products.

However, doing what I meant to do originally, I offer the following as a
challenge.

If the HP e3000 is going to stay around, I believe that Carly is partially
right when she emphasizes an openness what customers want and need.  She is
also correct that customers want solutions, not just platforms.  I think
that if customers are truly provided solutions that interact, that handle
their business needs, and where they are not locked into one solution (but
one vendor), then they won't care what platform the box runs on.

To answer Carly's challenge regarding open systems, the 3000 has been touted
by some (including myself) as an open platform.  That's where the rub comes
in... how have we defined openness with a proprietary system like MPE/iX?
We have, at various times, defined it as...

* being able to be used as a full web server?  The answer here should be
obviously be yes... but will it scale large enough to truly function in that
capacity for most business?  That depends on what they are trying to do with
it.  A large-end 3000 might work for the needs of most traditional
companies, but it may a questionable solution for "e-companies".  But, this
isn't "open", it's "Internet ready".

* being able to be used as an internal network server communicating with all
forms & types of hardware on the network?  Again, I think the answer is yes,
but is the system security configurations built to handle truly using it as
a server?  I would say (based on the numbers of NT boxes even in HP e3000
shops) that the answer is probably no for larger installations.  But,
again... this doesn't define a system as open.

* capable of running applications for all aspects of a business?  With the
limited number of applications choices available, the answer would be a
resounding no for most businesses.  Truly there are niche applications and
some vertical markets where this may be true, but there is a noticeable lack
of adequate applications.

* capable of running virtually any application?  Well, that depends.  Yes,
it can run virtually any type of application - but, again, in many areas
there just aren't applications available.  In addition, the industry
standard applications (SAP and Oracle, for instance) simply do not run on
the box.

* utilizing an open-source concept for utilities and system software?  No
yet... but we are getting there.

* hosting a variety of application options in a vertical market?  Yes, this
is really one of the biggest stumbling blocks.  Many of the vertical markets
that the 3000 does so well in have only one primary vendor - and that scares
potential customers off.

* being hardware vendor independent?  Absolutely not... and that's one of
the things that make it so good.  MPE is uniquely adapted to the PA-RISC
hardware.  It doesn't have to be able to run on every platform, so it can be
more finely tuned for the specific hardware.  [I think this is one area that
the open source and UNIX / Linux variants will find difficult to achieve -
true hardware / software tuning.]

Anyway, I think the point is made that despite the fact that MPE has Posix
and is web-enabled, it is not an open system.  Consequently, potential
customers must be sold on the concept of the proprietary system and the
advantages that such a coupling enable.  Now the question is how do we do
that?

* Excellent systems developed by knowledgeable, quality staff
* Educational encroachment so that kids can learn these things in the
university
* Multiple solution options
* Marketing... marketing... marketing.

Basically, the user community needs to address the first three of these, but
HP clearly needs to address the fourth option.  In the eyes of many, HP is
not doing their end... but we haven't been doing ours either.  So, maybe we
still need to call HP on the carpet about their marketing efforts (and the
visibility) of the 3000... but we better be very careful.  As a user
community we need to ensure that our developers are addressing the
additional items on the list.

So, now... it's up to us.  How do we do it?  Without whining, but
contstructively...

Glenn