Yosef writes:
> From what I have gleaned from what you wrote about the
> session HP gave a firm commitment, with caveats, to think about giving its
> blessing to the emulator project.

HP at this point appears to believe that an MPE Platform Emulator is
potentially the best solution for customers who need to run some or all of
their MPE code into the future so that they can do it on supported
*hardware* such as a (then-)modern multi-gigahertz Intel box.

A "Platform Emulator" (Mike Paivinen's term, which I like as it
differentiates our emulator concept from the migration vendors who are
currently using "emulator" to describe all sorts of things) is a SOFTWARE
PROGRAM that SIMULATES the HARDWARE of an HPe3000 computer system
sufficiently that it can BOOT, INSTALL, and RUN the UNMODIFIED MPE/iX
operating system.

As such it looks to the system it is running on as just one more application
program running on the system, but on the inside it is a true HPe3000,
capable of running existing MPE applications, etc.  An MPE application
running inside the emulator will not (easily) be able to tell that it is not
running on a "real" HPe3000 box.

HP has indicated that they have no interest in developing such a Platform
Emulator themselves, but that they will "get out of the way" by allowing
license transfers and new license creation for the MPE/iX operating system
that would allow a user to legally execute the software on a non-3000
hardware box.

Currently HP appears to be going down the path of taking a hands-off
approach, creating new MPE license terns that ignore the small complication
of how you make use of the software once you have licensed it.  This would
allow anyone/everyone to go write their own emulator if they want to.  This
has the advantage that emulator writers do not have to enter into a complex
agreement with HP, but it has the disadvantage that currently the emulator
makers would probably not be able to bundle the MPE software with the
emulator, requiring the user to acquire the software from HP independently.

Also without an agreement between HP and the emulator writer, there is no
justification for the emulator writer to enforce things like the "must run
on HP hardware" restriction, or to try to maintain the integrity of the
HPSUSAN and HPCPUNAME values seen inside the emulated environment.

Also the complication of having to potentially acquire the emulator and MPE
software separately and the unknown cost of a new "emulator" license for MPE
are likely to limit the chances of a commercial emulator product.

These last few issues are likely to result in further evolution of HP's
plans over the next few months.

Currently there are two companies who have expressed an interest in possibly
developing a commercial Platform Emulator program.  Ourselves (Allegro) and
SRI in Geneva Switzerland (who currently make a successful VAX emulator).
It is also possible, given the current HP licensing plans, that a "freeware"
or "Open Source" emulator could also be developed at some point.

It is certainly possible that the HP licensing costs/terms and restrictions
that they eventually settle on will result in both companies deciding that
there is not enough money to be made from the project, at which point you
are likely to get *no* commercial emulator.  At that point your options are
to wait for, or help develop, a freeware emulator (though whether this will
ever be of the quality needed to run your business is questionable) or you
can acquire the funding to pay one or both companies (or someone else) to
develop the emulator for you.

As far as the restriction of running MPE under a platform emulator only on
HP hardware, this was in fact *my* suggestion to HP in my original proposal
to them that they should give me the right to distribute MPE (essentially
for free) along with an emulator product.  The requirement to run on "boxes
that say HP on the front" (or Compaq now) would have provided some comfort
to HP that emulated MPE/iX hardware would not be taking sales away from HP,
at least through end-of-sales or end-of-support of the "real" MPE boxes.

> What is not clear is what emulator project encompasses.

I know exactly what it encompasses :-)  What questions do you have?

> I would like to suggest that the BOD ask these people and others to
> form a committee to define the emulator project.

Good software is rarely defined by committee.  A quality HPe3000 Platform
Emulator will only be developed by an individual or a small team of people
working together with adequate funding and lack of outside interference.

If you intend to write an emulator yourself, or wish to produce a
specification and then employ someone to produce it as a work-for-hire then,
by-all-means, use the design process that you feel is best suited to your
needs.

Gavin