So, what are you doing on this 2 year anniversary
of the fateful day of the HP 11/14/2001 announcement?

No, I don't mean what celebration, but what are you
working on or doing today that is an indicator of
what you have been led to explore over the last 2 years.

Interestingly for me, without even thinking about the
date, we happened to schedule a customer seminar today
that covers an introduction to PostgreSQL and discusses
the methods we are using to migrate TI databases.

I am however, wearing the HPe3000 dress shirt I got
from HP at HPWorld 2002.

Oh what an unfortunate reality - that the HPe3000 is now
relegated to 'shirt logo' status and not 'server sold' status.

But then I can do more things with my PostgreSQL databases
than I could ever do with my TI databases, all for a lot
less money. More bang for the buck, more value received,
not just perceived.

HP got it right both times. First, when they said
"customers buy solutions, not technology", and second
when they realized the HPe3000 was being trumped by less
expensive, more functional solutions and should be retired.

Lest you think I believe this was inevitable and a pure
result of market evolution, read on...

Unfortunately, HP didn't get it right in the first place,
when they missed the opportunity to completely redefine
proprietary business computing cost structures with an
introduction of the a/n class systems at performance and
price points that would make all CTO/CIO folks take notice
- and be required to consider in their buying decisions.
Looking back, the decisions appear to have been made for the
short-run and not for the long-run in mind. Almost a
self-fullfilling scenario. You keep the price points in
line with past practice because you believe the business
is not growing and you get a guaranteed 'no growth' result
because you don't stimulate enough additional buying opportunities.

But taking the chance would have been in the best interests of
the installed base customers for the long run. Especially since
the way it played out didn't generate much of a 'long run' anyway.

HP didn't get it right, but Stan got it right. When, at the
SIGSOFTVEND meeting where HP unvieled the soon to be sold
a/n boxes, after seeing the performance numbers he
remarked: "They have killed the 3000".

As we all know, you adapt or you die. HP didn't allow the
HPe3000 to adapt. And now it is dead.

And two final notes:

* HP hoped not allowing 7.0 to boot on 9x7 boxes would spur
  upgrades of 9x7 boxes to new a/n systems. A better idea
  would have been to sell systems so dang fast and cheap that
  anyone with a 9x7 box would be happy to upgrade.

* At that SIGSOFTVEND meeting HP marketing actually said they
  were expecting customers to buy 'a' class boxes for web servers.
  These would be the same systems that, while advertised as 110mhz
  and 150mhz systems, actually performed more like 55mhz/80mhz
  and run an o/s that has one of the slowest tcp/ip stacks in production
  and include the cost of a database whether you needed it or not.
  No, you don't buy the hpe3000 to be a web server. What you do is
  you hire marketing people who know your products, the industry
  products, and what people care about when they make buying decisions.

duane

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