Re: Market Driven was RE: Binding a type to a union.

[Robert Dewar <robert_dewar@my-deja.com>]

In article <NBBBJNOMKDIAJALCEFIJMEMODFAA.rleif@rleif.com>,
  comp.lang.ada@ada.eu.org wrote:
> From: Bob Leif
> To: Robert Dewar et al.
>
> Robert Dewar wrote, "we are certainly not in the mode of
> asking the community for neat ideas!'
>
> This process is called being market driven. You have just
provided a very
> good explanation for a significant part of Ada's lack of
commercial success.
> Most successful companies know the value of consumer input.

Absolutely, we definitely know the value of consumer input, it
is just that we do not regard miscellenous suggestions on CLA
as consumer input. For us consumer input is carefully thought
out suggestions from consumers in our target market. Read my
previous post in another thread about how everyone is sure
that XXX is the key to Ada success. XXX is different for many
contributors to CLA, but generally we simply don't find these
suggestions to be very valuable as consumer input.

Oh yes, and it is typically the case that the people who are
sure that XXX is the key, are also sure that failure to
implement XXX is what contributes to Ada's commercial success.

Well so far, Ada Core Technologies is being quite successful
commercially and otherwise. We have made some missteps, which
interestingly are cases where there was a loud constituency for
a particular port, notable cases are the Mac, which many people
were sure had a large crowd of people just itching to get their
hands on a Mac Ada 95 compiler [actual experience, we had one
customer who tried to get MachTen to work, but had insuperable
difficulties using it in a production environment], and another
notable case was the DOS port, for which there was never even
one customer (so yes, Vladimir, we do have some cases where we
invest our own resources in failures :-)

Setting priorities and figuring out where to put resources
and what new features are needed and valuable, is not at all
an easy process to be sure. Part of the trouble is that many
people making suggestions do not have a good overview of the
Ada market, but rather tend to see just one little part of the
market, or concentrate on one possible extension of this market.

We are certainly working on new features for GNAT all the time.
If we are not working on your pet feature, it means that either
we don't deem it valuable, or we see other things as more
important given our estimation of consumer input, and
incidentally not just consumer input, but consumer needs.

The static elaboration is a good example. Lots of Ada users
have a lot of trouble with elaboration, but I would not expect
consumer input to say "hey, how about implementing a static
mechanism for elaboration which bla bla bla....". Instead this
is a case where we perceived a need, invested considerable
resources in providing it, and sure enough it has proved
invaluable to many users (some know it, others don't even
know it, they just don't run into troubles which they otherwise
might have :-)

Robert Dewar
Ada Core Technologies


Sent via Deja.com http://www.deja.com/
Before you buy.