Ron writes:

> At the NewsWire we'd like to see an MPE-Image transfer to an
>  organization dedicated to the environment's growth become a part of
>  the virtual CSY's to-do lists. Holding onto a discontinued product --
>  for the express reason that "we don't think anybody can do it as
>  well" -- fails to serve the needs of the customers who are not moving
>  away. I leave it as a study of business strategy for others to
>  speculate why HP's holding onto MPE-Image might be otherwise so
>  important.

Ron's statement, "At the NewsWire we'd like to see an MPE-Image transfer to
an organization dedicated to the environment's growth" worries me a great
deal. There are several ways to interpret that sentence.

One interpretation is that HP should essentially just give MPE to one of the
current vendors. I am wholly against that option. It favors that one vendor
to the exclusion of everyone else and will in the end probably be the
shortest path to the demise of MPE, excluding the current road we're on, of
course.

A second interpretation is a current vendor buys MPE from HP at some
significant cost. If nothing else, this seems more fair to me. The purchasing
organization would have a substantial motivation to see MPE succeed, but my
basic concerns still persist. Ultimately, I think that MPE will fail in this
model as well, even though there will be a short period of time when MPE is
actively and aggressively advertised, an activity different than we're used
to seeing.

The third interpretation -- and the one that I don't think Ron is advocating
-- is by far and away my preferred path: that all of the source and the
documentation for MPE be given to at least three, four or five different
organizations, each to be an independent but interlocking curator of the
code. While this system would not be "open source" per se, it would offer
both the stability of independent repositories, making MPE essentially
unkillable, and an agreed-upon framework to carry MPE forward into the
future, allowing modifications to be made in a controlled environment. But
most importantly, it would allow inexpensive licences (in the range of $150)
to be put out to the world at large.

We are at that point that if MPE is to survive, we must consider that we are
at that time in the evolution of MPE that we are at a complete restart, where
we begin again from ground zero. The large enterprise customers that HP found
so attractive from a cost-of-sales point of view are simply gone. They do not
now consider MPE as a reasonable alternative, nor will they for a great time
to come. If MPE is to expand, to become as popular as we all believe it can
be, the future of MPE lies in the small, cost-sensitive business customer who
would very much like to run MPE (and its IMAGE database) on an inexpensive,
high-speed Dell box. To this customer, an MPE distribution structured around
a very near-Red Hat-like pricing and distribution model offers a very
attractive alternative to Linux.

But even more than that, this is the only model that will get MPE back into
universities, colleges and trade schools, which are MPE's most important
venues. But as we all know, if MPE succeeds here, in the classroom and in the
small business market, it can grow once again to be an enterprise class
solution.

I simply don't believe that any other path to success exists. Evolution is
not random. Rather, it's chaotic, where the initial conditions essentially
predetermine the outcome, and small changes in those initial conditions
portend significantly different outcomes, most of which end in failure.

Wirt Atmar

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