>>
>><<My name is Kriss Rant, and I work in product marketing for the HP 3000.
>>
>>We would like to understand your future business requirements in the area
>>of application and database environments.
>>
>>Our objective is to ensure that customers who are using common
>>application development products--IMAGE/SQL, ALLBASE/SQL, KSAM, COBOL,
>>and VPlus have reasonable choices for moving forward.>>
>>

Here's a radical suggestion.

Assuming that HP are going nowhere with IMAGE/SQL - big assumption, and
thet haven't said that, remember, but assuming - and assuming that
there's a UNIX port somewhere - another big assumption, but we all know
the rumours - why don't HP put it in the public domain, rather than just
let it wither on the vine?

Advantages:

UNIX gets a 'lowest common denominator' database (in the nicest possible
way) that everyone can write for without wondering if the client has it
or not. (Just like we could on the HP3000 - remember?).

Us HP3000ers all get guaranteed jobs for life because we know this stuff
- helps ease the pain if the HP3000 becomes 'functionally stabilised'
(i.e. tag-on-toe time). Of course, HP *could* make a small design change
to stop there being migrating secondaries, before releasing it, but why
steal our advantage?

And maybe, just maybe, people who liked this database would even start
looking for the platform that it really screams on....

Disadvantages:

'It doesn't run fast on UNIX'. So what big DB does? So throw hardware at
it. So who will notice just how fast the database is or isn't under
those multitudinous layers of middleware, screenware clientware and
appware?

'It isn't relational'. An excellent differentiator (see next objection).
No, it's a networked database, in the very very old meaning of that term
(but watch how the connotation would change).

'It will upset HP partners like Oracle'. What? You mean they couldn't
pitch their modern, supported, relational, distributable, tool-enhanced
products against an aging, unsupported, networked.. (well, you get the
picture)

'It isn't open' So throw it open. Call it OpenNetDB. Give it to a
standards committee. Whatever. If it even matters.

Here's a heresy(?). ** I don't think anybody much cares about 'Open',
per se, really **. What people want is not to be cut off by a hardware
choice. Unix doesn't sell over MPE/iX because one is 'open' and one is
'proprietary'. It's not some doctrinaire thing about who makes the
rules. It sells because if you buy the Unix box, you cut yourself off
less from future potential apps. Hard, but there it is.

Just watch the rush to NT.If pretty much all the software that people
want is available on NT, as it will be, they won't care that NT is
Microsoft's, and Unix is run by standards committees.

Actually, I think they might prefer NT, as we now might prefer MPE/iX,
and for some of the same reasons.
(i) the rules might be arbitrary, and it might be hard to get to
whoever's making them with your bit of special pleading, but at least
there ARE rules, and we all know what they are.

(ii) MPE/iX and NT are much better optimised for mission-critical
business processing. People use Unix because it's got the apps, not
because it's the best OS for the job.

Image on NT? Now there's a thought....

Roy Brown
Roy Brown                    Phone : (01684) 291710     Fax : (01684) 291712
Affirm Ltd                   Email : [log in to unmask]
The Great Barn, Mill Street  'Have nothing on your systems that you do not
TEWKESBURY GL20 5SB (UK)      know to be useful, or believe to be beautiful.'