Re: Ada 95 for an ARM-based bare board?

[dewar@gnat.com (Robert Dewar)]

"Juha Valimaki" <juha.valimaki@xdsloy.com> wrote in message news:<akf4bg$454$1@phys-news1.kolumbus.fi>...

> If some kind of proof is needed why small markets _need_ every customer look
> at it this way:
> X is number of people using free tools
> Y is number of people using low end commercial tools
> Z is number of people using high end commercial tools
> 
> X+Y+Z is the group of all Ada users. If you take away the free tools then N%
> of group X will buy a tool. N%*X+Y+Z > Y+Z, if N>0. Note however: Y+Z is
> likely to be smaller than X+Y+Z, but the market will be larger in terms of
> money!.

A completely bogus argument, because it ignores the fact that the
existence
of a free tool may increase the total number of users, so the X/Y/Z
are by
no means constant, and you have no evidence that they are (or rather
would be)


> For me it seems you are afraid that low end solutions would eat up your
> revenue

Nope, that has nothing to do with it, it is just that this end of the
market
is not what we specialize in, and we believe our success (our sales
are substantially up this year) is significantly due to the fact that
we keep
a narrow well defined focus.

> and that Ada would be used more in general purpose programs that
> don't focus on mission critical code

No, we don't believe that at all! If you believe it why not start a
business
aiming at these users, since it sounds like you believe it would be
profitable.
By the way, notice that you are also agreeing that the X/Y/Z may not
be
constant in your above equation (though in the opposite direction from
me :-)

> That's a perfectly legal reason to suppress competition :-)

Nobody is suppressing competition here. We are one of the very few
software
companies that invites you to compete its market, and hands you at no
cost
the technology that you need to compete.

> I really believe Ada tool vendors would understand that there has to be some
> way for people to try Ada and do research without large cost. 

For trying ada and doing research, the public version of GNAT is
definitely
appropriate, and for sure does not have a large cost.


> but I
> think many small companies would easily pay $2000-$4000 per developer (PER
> DEVELOPER is the key here).

Well of course for small teams of people from 3 on up, that's
precisely our
price range, and you are right, many small companies are prepared to
pay this
for high level support.

The one case where we do not cater to this model is for smaller groups
(of
1 or 2 people), but as I said earlier, we did this at first, and found
that
these isolated users of GNAT expected far far more support than larger
groups
and we simply could not afford to give them the high quality support
that we
guarantee at the single seat price. We actually found that the typical
pattern
was that one person teams asked far more questions than a five person
team,
but we really can't charge more, but we do charge the same!

Robert Dewar
Ada Core Technologies