I have always had the opinion that HP's, and some vendors, tiered pricing
model was flawed.  I think this contributed to (but was not solely
responsible for) the demise of the HP3000.

In particular, I am referring to the concept of defining tiers based on CPU
speed as opposed to the number of users (which is a more modern concept and
a better measurement of benefit).  The flaw of associating "benefit" based
on CPU speed harks back to the early days of tiering on the classic 3000's
when there was no control or license for the number of users.  Indeed, user
count was usually limited by the speed of the particular model and so the
connotation between these metrics seemed more appropriate.  Until sometime
in the mid-1980's HP simply licensed software to be run on a particular
HP3000, prior to tiering.

Since the introduction of MPE/XL this relationship between speed and users
began to fall apart.  Unfortunately, some vendors (HP included) did not
always adjust their tiering model accordingly.  (HP did correct this
problem with the introduction of the "A" and "N" series, but this was too
little, too late.)  Consequently, a smaller company who could perhaps
afford a faster CPU, would have to forgo the option because the
consequential software upgrade pricing penalty was prohibitive.

Performance improvements perceived by users because of faster response time
or throughput only benefits the vendor with increased user satisfaction.
Assuming the total number of users does not change, there is no increase in
vendor costs because an application runs faster than before.  Therefore,
there should be no additional fees for a license transfer, other than a
nominal license transfer fee.  If I just want my processing to take place
in a shorter time frame and can afford the hardware, this should not be of
concern to the vendor.  This would no doubt have resulted in more CPU
upgrades for HP because there would not have been a (sometimes large)
software penalty to pay.  Because some vendors just copied HP's flawed tier
pricing model the problem multiplied.  Paying more for CPU speed makes no
more sense than pricing the application based on how much profit a user can
derive from using it.

I believe that many more HP3000's would have been sold if this artificial
and difficult to justify barrier had not been in place.  Surely vendors
would gain more revenue from a larger volume of installations of any size,
than they would from extracting premiums from those few willing or able to
pay it.  It is short term and dead-end thinking to impose barriers between
a vendor and the potential marketplace with one-sided pricing models.  I
believe this was a factor in the ebb of the "ecosystem" surrounding the
HP3000.

John R. Wolff
Vice President/CIO
LAACO, Ltd.

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