On Wed, 15 May 2002 22:46:24 -0700, Bill Brandt <[log in to unmask]> wrote:

>I'll be the first to admit that I don't read all the digests I get - they
>are huge - but the immediate thought that came to me is why would anyone
>want to go to the expense to migrate from an HP3000 to yet another
>proprietary system?

Answer: Perhaps they are looking for stability?

It depends on the objectives and perspectives of each user, of course.
Some will be development oriented and have customer solutions for their
businesses that may actually be superior to any available purchased
packages.  Others will be just the opposite where a package that runs on
several platforms will be the best answer, even a proprietary platform.  I
think that what "camp" you fall into has a lot to do with how independent
you feel with your particular application, and less to do with popular
trends.

It is interesting that you have sympathies for the popular platform theory,
but do not believe in it quite strongly enough to take the risk (or
expense) of migrating to it from the HP3000.  Apparently, staying on the
HP3000 is the best answer for you (so far); i.e., you have more to lose
than to gain by migrating.  I feel the same way in our case.

The proprietary platforms (MPE, OS/400, VMS to name a few) have the longest
track records of stability and continuity over the more popular platforms,
such as UNIX and Windows.  UNIX (with its many flavors) is a dressed up
1960's design that gestated in the university environment for decades
before escaping into the outside world, and has for whatever reasons become
popular in the late 1990's.  Windows (with its many flavors) has been, and
remains, a work in progress over the last 10 years and could be called a de-
facto standard (that changes a lot).  Users have no "right of
compatibility" or control over the effects these changing systems will have
on their custom applications.  The so-called standards are even subject to
change by many outside forces that will effect your business, but may not
contribute anything to it but added expense.

UNIX (including Linux) has made many attempts at standards within an
overall non-standard, but always seems to miss the mark and is still
undergoing many changes.  Windows is constantly evolving, if for no other
reason than to keep Microsoft cash flow high.  From the custom application
users perspective (at least this one) these two platforms have a rather
unstable track record without promise of compatibility and present a
constant challenge to stay in step.  It is challenge enough to keep custom
applications current with a particular business without the added burden of
trying to stay in step with "dancing" operating systems.

Users that use packaged solutions have off-loaded much of the OS problem to
their vendors to worry about.  In addition, they may have options as to the
hardware vendor of choice.  There is nothing wrong with this strategy, if
it works for your business.

On the subject of homesteading with MPE I would like to make the following
observation.  If it were not for the fact that both MPE and the HP3000
hardware it runs on are so stable, homesteading would be a tough choice
indeed.  But, given these blessings and with careful planning and the
support available in the community, this choice is feasible for the next 10
years.  One of the difficult realities HP has had to face is this very
fact  --  they built too good a product and many customers/users were not
dependent upon them.

In the end, these considerations are really just business decisions.  My
opinions and observations are taken from the standpoint of a business
manager, not a programmer.  So, to show how viewpoints can differ (perhaps
both being "correct") the question that comes to my mind is:

Why would anyone with a working custom application want to go to the
expense to migrate from a stable HP3000 to the popular, but unstable, UNIX
or Windows platforms?

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