While I often agree with Duane's opinions, I'll have to disagree on this
one:
> 2. Those who think HP is making a mistake by focusing on
> commodity systems
>    are missing the point entirely. There is nothing wrong with being
>    in a commodity business. Does anyone here have a problem
> with Proctor
>    and Gamble?  Nestle? Or any other company selling
> commodity products?

P&G is not just a commodity company.  If you go to their web site, you will
see plenty of products that serve a premium clientele.  (IAMS pet food,
salon class shampoos and conditioners, and even high end personal cleansers
like Oil of Olay.)  Similarly Nestle owns Perrier and along with regular
Friskies cat food, one can buy the "upscale" Fancy Feast version.

To me, the secret of these "commodity" companies is that while they serve
the commodity market, they are always "Inventing" new ways to market premium
products using their current assets.  (How far away is Oil of Olay soap made
from one of the cheaper brands?  A few yards?)

>    It has been proven that if you are a growth company and
> want to provide
>    continue shareholder value you need to own markets to be
> successful. If
>    you don't own a market then you always fall behind.

What about Apple?  They certainly don't own the market.  Are they
successful?  I am sure quite a few are surprised they are still around.
What keeps them around?  Innovation.  Commodity company?  No, but they
market a product in the midst of a commodity environment.

I think MPE's fate was sealed in just one stretch of short-sited management.
I thought that they had chosen the convergence path.  They started sharing
the hardware with the UX group.  They added the HFS to MPE.  The POSIX shell
was softening up some of the MPE developers and created an environment that
was comfortable to non-MPE developers.  What I thought was going to happen
next was a hardware or software bridge between the two.  With
multi-processor machines, how hard would it have been to add an MPE process
that ran on a dedicated processor card under UX?  How hard would it have
been to create an MPE emulation layer under UX?  Maybe less than trivial in
both cases.  But look what possibilities such a path would have created.
You have reduced your manufacturing to one box, a smaller team to maintain
MPE (hardware drivers are now shared effort), kept your current customers
happy and open the possibility of MPE users buying products that run on UX.
This makes the developers life easier because now their customers already
have a UX box.  They could have developed with confidence but instead we
have an ecosystem that is searching for food in the IBM pool.  (Who BTW took
the approach of convergence and is not going to just EOL the AS400, they're
going to make it difficult to move to another platform by making it easy to
stay on the next generation of IBM systems.)  Is it too late for HP?  I
don't think so.  There are already companies who think this is worth it, why
not partner and reduce the acrimony of the client base?

But I have to agree with Duane that the world is moving forward and we must
evolve to move along with it because HP is certainly not making the case for
intelligent design...

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