Jon writes:
> [...]I'd like to hear the thoughts of the three possible PA-RISC
> emulator vendors (Allegro, SRI, and Strobe Data) on pricing
> structure.

Ok, if we presume for a moment that getting a copy of MPE will cost Jeff's
(currently mythical) $500, then that puts an absolute lower limit on the
total cost of an emulator to the end user of $500 even if we give the
emulator away.

It might be less if you have a "real" 3000 license to sacrifice, but even if
the transfer-fee is like $100, I don't think that changes the dynamics
dramatically as the value of even a 9x7 is like $100 or more these days so
the total cost is probably similar.

It has always been one of my goals to let everyone who wants MPE on their
laptop to be able to have MPE on their laptop, via some sort of free/cheap
license for a version of the emulator which doesn't cannibalize sales of the
emulator for "production" use.

The advantages to me in providing a cheap/free version are that it satisfies
the community desire for such a thing among people who are unlikely to pay
thousands of dollars to run MPE slowly on a laptop, that it gets our product
into the hands of people who may later decide that this is a good solution
for those bits of their computing environment that they would like to leave
in production on MPE "forever", and that it gives me a larger customer base
from which to draw feedback, enhancement requests, etc.

So the question then arises, what would be an acceptable way to distinguish
a cheap/free version of the product from the "full featured" version, and
what, if anything, would people be willing to pay for it?

Some things we could do to differentiate two "tiers" of emulator product:

Restrict, by license, the use of the product for "production use" for some
definition of "production".  Of course this puts a lot of trust in the end
user.  My definition of "non-production" would likely allow development and
support activities since I don't believe there actually are any real
"hobbyists" out there for the most part.

Limit the performance of the cheap version.  This seems like a dead-end
because we'll already have a lot of work to do to get the performance up to
being even minimally acceptable for production use.

Limit the resource available in the cheap version.  We could limit you to
1GB of disk space, or 128MB of memory etc.

We could limit I/O connectivity in various ways, or make all more exotic I/O
connectivity come at an additional cost.

We could set the HPUSERLIMIT to, say, two.

Limit (or eliminate) network functionality to some degree.  No incoming VT
or telnet sessions for example (or only one at a time).

Fix the HPSUSAN number and/or HPCPUNAME at a fixed value that says "this is
a non-production version" for which 3rd party vendors might not give you a
production license code.

If you have to pay $500 for the MPE license, can you still afford a few
hundred more to get a more professionally packaged emulator with some degree
of support, or would you rather have a free or almost-free version that you
download and which we can't afford to even answer your emailed questions
about :-)

We could not sell a cheap version independently, but include a certain
number of limited development/support licenses (limited as above) with each
"production" license.  So to get your laptop 3000 you would have to get your
company to buy a production license.

We could try to do more "tiered" pricing (ick) on the production version
such that we try to extract from you exactly the value you're getting, based
on number/speed of CPUs, the received effective MPE CPU speed, etc.  MPE
already has this convenient mechanism for crippling itself to whatever
degree that you want :-)  In this scenario the "918" (930? :-)) version of
the emulator might be so cheap that you wouldn't need more than one
"version".

Anyhow, there are some topics for discussion.  I'm interested in hearing
from people who think they are potential customers of any kind of platform
emulator, and what you think of these various ideas.  Of course if you want
to tell me exactly how much money you have to spend, that would be fine :-)

You should consider that the market for an MPE platform emulator is limited
at best, even without two or three competing products in the space, and that
one or more products will only get developed if the respective company
thinks that it will at least not lose its shirt in the process.  The more
serious danger is that they'll think of something that's simply better to
spend their time on (from an ROI point of view) since the MPE platform
emulator is a relatively high investment for relatively little return.

I would also be interested in hearing what people would guess is the number
of production emulator licenses that could be sold.  I think you'll find
that the actual number is actually around 1/10 of what the average guess
would be.  Customers seem to always over-estimate the size of markets (and
the sizes of the companies filling those markets for that matter).

G.