Re: The great Java showcase (re: 2nd historic mistake)

[dewar@merv.cs.nyu.edu (Robert Dewar)]

Wes says

<<Yes, if a product (Beta, Edsel, whatever) is not what people want,
then you can certainly argue that it's not superior.  But when a
product is _exactly_ what they claim to want, yet they won't buy
it, what can you do then?>>


listen harder!!

This must mean that they have some requirements that they are not
articulating clearly. For example, they prefer software written by
a big company like Microsoft, or they feel more comfortable using
tools that everyone else uses. These are requirements that are
just as important to people as simply looking at technical aspects.

Why do people pay Bayer large amounts of money for simple
chemicals that are available much cheaper from others -- well
they feel more comfortable that they are really getting what
they want with the Bayer name behind it, and of course Bayer
spends large amounts of money convincing people that they feel
this way (which does not necessarily mean that it is in an
inappropriate attitude).

Yes, it's sometimes frustrating that customers seem to make "illogical"
decisions. We often see people using Ada technologies that purely in
technical terms seem clearly inferior to GNAT by any objective analysis.
But there are many reasons for this. Often it is simply a matter of
personal relationships, if you have dealt with person X for a long
time, and trust them, and they have not let you down in the past, then
it's not at all unreasonable to depend more on that trust than on the
current technical state of tools.

Being in business is all about understanding how all these factors work
together and understanding not just the technical needs of customers,
but their complete set of requirements, and identifying the cases where
you can meet that complete set of requirements. We often end up 
suggesting people look at some other Ada technology if we feel that,
considering all the requirements in this general sense, the customer
might be better off with some non-GNAT solution. Of course, being a
service and support company, rather than a "heres-the-shrink-wrapped-
software-take-it-and-pay-for-it" business, we probably work harder to
make sure that we *do* meet all the needs of customers, since we need
them to feel they are getting value for their support contracts!

Anyway, as I say, I understand the frustration, many projects that could
be more successful if written in Ada are not, and those of us who know
that spend a lot of time asking ourselves how that can be fixed. Well
the answer is that it cannot be fixed generally, and the important thing
is to concentrate on an incremental approach, where you make sure that
current use of Ada *is* successful, and rather gradually increase the
awareness of these successes. I don't see any other approach that will
work!

Robert Dewar
Ada Core Technologies