OK, there are a few comments that I would like to add to Douglas' post.

Here we go :

> HP has done the citizens of the United States a great disservice by
> discontinuing the HP3000 and MPE:

Not only citizens of the USA. Did you know that a customer of mine, actually
a workplace medical  association in France, runs all its software on an
HP3000 ? Another customer of mine in Paris is a health insurance firm, I
think what you call an HMO in the States.

So I would surely not restrict the scope of OpenMPE to the States, nor the
influence of the Nov. 14th announcement.

> There are many Federal,
> State, and Local
> Governments which use the system, and, by discontinuing the
> product line,
> has imposed upon the citizenry of this Country a migration in
> a time of
> shrinking budgets, such that we ALL pay.

At the end of the day, that's sure. All migration, consultancy services, new
hardware, new software and other associated costs will be passed on to
customers (price increses) or, in the case of services, onto the taxpayer's
budget. That, too, is true all over the world.

> Traditionally, governmental agencies will not fund a private
> concern for
> the sake of a community of for-profit industries, and, so, asking a
> governmental agency for even $1,000 per year, let alone
> $2,500 per year, is
> out of the question.

In France, public agencie, administrations and other governmental entities
have very strict procedures to commit money, that begin with a request for
submitting offer. But these entities cannot be expected to issue a request
when they're not sure what's coming forward.

> The next area of concern is management of an independent
> CSY—particularly,
> as I understand it—about half of the CSY personnel are in
> India and the
> other in the United States, with a few in various other parts
> of the world;
> if this be so, how will the infrastructure be administered?
> How will the
> people be managed? How will the goals be set and performance measured?

It's (my guess ...) up to CSY to answer. IMHO, any entity or corporation can
have multinational offices. For instance, APPIC has 6 entities spread all
over the world. So whether CSY ownership is somehow transferred off of HP,
there islittle change to expect, it will just be another multinational
corporation, with a branch in, say, India.

Goals, objectives and performance can be measured accross national
boundaries. Whatever the ownership of the entity, a performance evaluation
is a performance evaluation. This is a no-brainer.

> For example:  It took the CSY seven months to fix certain
> problems I had
> discovered while experimenting and trying to get the
> Apache-Java server
> going, [and in the final analysis it was because we did not
> license the DS
> component of the operating system]; it was clear that the intended
> functions we needed SHOULD HAVE been embedded in the FOS, but
> they were
> not, and, so, we got the DSLINE functionality for essentially
> the price of
> maintenance.

This is more tricky. It raises a few issues :

1) How to keep needed expertise within a new entity (New CSY ? OpenMPE ?
Whatever ...)
2) What will the support model be ? Should we emulate HP's WTEC model
("Sun-never-sets" response centers taking over time zone after time zone)
3) Are we talking of just FOS or the whole product line ? Do we include
subsystems like Allbase ? TurboStore ? Disk Mirroring ? Compilers ? Etc ?

> If the CSY claims that it would take 100 person years for a
> minimum of six
> elapsed years to port to IA64, then what confidence would we
> have their
> attempting to port MPE to anything beyond the current RISC
> environment?

Back to the core issue : once PA-RISC is obsoleted, what if we don't have a
viable IA-64 version of MPE ?

> Then there are questions of management of things like
> DTCs—now there is a
> nightmare for most of us—though it might be simple to
> implement a terminal
> server environment, it is yet another subsystem which needs
> considerable
> thought before any migration.

The complete product line is relatively vast, and I expect that choices will
have to be made as to what should be migrated berfore, and what should be
migrated later, and what should not be migrated at all. Call that
prioritizing issues relatively to a constrained budget and customer/user
demand. This is exactly what was done when MPE was migrated from the Classic
architecture tp PA-RISC.

Just a  question : should the DTC be migrated at all ?

> Then there is the question of emulating the emulated
> compatibility mode:
> Think SLOW.

Yes, once again exactly what happened in the good old times of HPE - MPE/XL
preliminary versions, when lots of code were still in CM. Still, we're
talking about running code that was originally designed for, say, 70s, that
were running at frequencies in the megahertz range, on IA-64 chips that run
at least at 750 MHz. Resulting performance may not be as fast as full IA-64
native mode, but it should still be "sufficient".

> Then there are migrations of people:  Some of the best of CSY
> have no doubt
> been looking for employment elsewhere, and some may have already been
> leaving; others may have already been layed off—in any event,
> since the
> current funding model is not from MAINTENANCE, but rather
> from NEW SALES,
> the absolute life of CSY is little more than two years, in a practical
> sense; is there a viable way to insure keeping a critical mass of the
> knowledgeable people together for the next two years?

This raises two important issues :

1) How to keep the needed expertise ? Or, if it has already left, how to
attract it back ? A few days back, someone else raised the issue of paying
enough enginners enough money to keep them in, or woo them back. That means
a salary mass of enough dollars, that has to be matched by enough revenue.

2) Since we are talking of mostly pursuing with a current installed base (I
have not seen much talk about new business ...), the revenue stream should
not be too much focused around sales (especially if we go toward a zero-cost
licence model !!!), but around support. This has the advantage of ensuring
an almost constant revenue stream ... providing customers are satisfied and
convinced they ought to remain on the platform !!!

> The best of all worlds would be an emulator board to run MPE
> on a Wintel
> box, just as some of the IBM Zos – OS/390 processors do; but
> what are the
> chances of that happening?

Not sure whether this is an appropriate track to pursue. If I get it, this
requires developing more hardware. This requires a lab with tremendous
expertise, then a plant to manufacture the hardware (or an outside
contractor), a supply chain to ensure deliveries, hardware maintenance, etc.
The software emulation (which may or may not include Linux, which may or may
not include an IA-64 port) seems to me a lot easier. And cheaper.

> Who decides the direction of MPE after HP?

That's if there is an "after", which is far from granted at this point.

> The County has already mapped
> migration to Sybase on Sun Solaris using PowerBuilder / Cold
> Fusion and you
> can lay good odds that other governmental agencies are
> already planning to
> migrate, if not already in progress as we are.

It does not hurt to pursue two tracks. That can be considered anticipating
one's future. At some point, considering the current situation (not past or
future situations), a choice can be made by the customer.

The partner/software supplier/consultant/integrator way of thinking has to
be different. It should include more parameters (revenue stream,
profitability, ROI, technical hurdles to be overcome, staffing,
multi-national issues).

> It is interesting to watch the politics within HP,

Yes, but for different reasons. I don't see a reversal of the 5-year plan
coming. I do not count on that. Even if the merger is voted down, even if
Carly resigns, even if she is thrown out ... IMO, these are two different
topics, and I address them separately.

> the Year 2001
> will be one that we all wish we could forget.

No. Speaking of 2001, what do we have ?

- Sept. 11 : We all would like to say "never again". But if we don't want to
see Sept. 11 again, we need to keep in mind what happened on Sept. 11.
That's pretty much similar to saying that if we don't want to see the
Holocauts again, we should never forget there was an Holocaust in the first
place.

- There is the HPe3000 5-year plan : This raises opportunities to "preach
outside the choir" the cause of MPE. The idea of spinning CSY off of HP has
been raised time and again by Wirt, and others. This Nov. 14th announcement
may be the jolt that was needed to either spin off CSY entirely, or
outsource (or otherwise transition) MPE outside CSY. This is a shock, but
this is also an opportunity. It should not be merely taken as a menace. This
is similar to what happened in the 60s, when most movie studios saw TV as a
competition and the only one who saw it as an opportunity was Disney. Of
course, Disney made an awful lot of money making stuff for TV, and
eventually all majors followed them and worked for TV as well.

I would not say I like what HP announced on Nov. 14th, but we have to live
on with that. So, instead of mourning (you mourn only dead people and
things, OK ?), we should try to understand the positive aspects of the
situation, and develop a comprehensive plan around these.

> What confidence could anyone put on the future of the group
> under these
> circumstances?

Totally depends on the group's approach. If we indulge in bashing (HP
bashing, Interex bashing, each other's bashing, whatever ...), we lose
whatever little credibility we may have gathered, and confidence can safely
expected to be nil. If we develop and implement a positive and constructive
approach, we reinforce the group's credibility, and may gather and keep
customers' (and HP's !!!) confidence in the future.

Oh, my, speaking for myself once again....

Christian "writes long mails, a la Wirt" Lheureux

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

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