Oh, where to start?!
I made a short presentation at SIGAda on this very
topic. Maybe I should turn it into a white paper
of some sort.
> I work for a major defense contractor at a site where we've done
> lots of work in Ada, and lots of work in C/C++. Recently, some
> influential colleagues have expressed concern over the viability of
> Ada. One of them has asserted that it would be "looney" for a new
> project to choose Ada, because of the limited availability of
> compilers and other tools, especially for new processors, and
> because of risks related to future availability of tools. This is
> from someone who is quite willing to acknowledge the technical
> advantages of Ada.
I'd suggest first looking at the products and the suppliers
that match your planned architecture. There are lots of new
Ada compilers out there for most major architectures. If
the products look good and the suppliers seem reliable, I
don't see the risk for Ada to be much different from
anything else. Most of the active suppliers have been
in business for quite awhile and don't seem to be going
anywhere. Just because a supplier provides C or Java
doesn't mean they are particulary stable long-term bets.
Look at the problems Borland had, for one example.
The risk of the language itself dying (as opposed to
any particular supplier) is nil. Like COBOL and FORTRAN,
it's been used too long by too many people to die. Also,
like C/C++, freeware versions are extant, and use in
universities is widespread, assuring long-term availability.
> Frankly, I think the concern is exaggerated, but it is obvious that
> the current market for Ada is much smaller than for C++, there are
> more compilers and other tools for C++ (or Java) than for Ada, and
> there is more vendor investment in C++ and Java than in Ada. My
Certainly true, particularly in areas like consumer-level
applications, C/C++ and Java (and for that matter, Visual
Basic) are ubiquitous. If the raw number of compilers
sold is the major criterion, then the question is already
answered. On the other hand, if one considers other
aspects, such as the complexity and quality of the
applications built, then it's worth looking further.
What I find odd is that languages like Eiffel and
Smalltalk are considered "cool", while Ada is often
stigmatized as "dead". This, despite the plain fact
that basically nobody uses Eiffel, Smalltalk, etc.,
for real applications whereas Ada is used quite a
lot on some very major programs.
> position is that there are good compilers and tools available, but
> I could use some data (opinions will be of little use) that
> supports the claim that the Ada market is not disappearing. I have
> used Dick Reid's data on languages in CS1 courses (posted here a
> few weeks ago) to show that Ada use is steady (and nearly matches
> C++) in this area, but I could use some hard data for the current
> Ada commercial market. Can anyone point me to such data? Is there
> any interesting news along these lines from last week's SIGAda
> conference? (Statistics from the PAL or Web sites such as Ada Home
> are interesting, but I suspect that information about where money
> is being spent would be more persuasive.)
So the point of my presentation was that the answer to
the question "Is Ada a Commercial Success?" is very much
dependent on the definition of "commercial". If you
mean video games and office suites, Ada is essentially
unused. I prefer to think of "commercial" as something
closer to "not military", to encompass application
areas such as commercial avionics, rail, power plant
controls, robotics, industrial process control, space
applications, image analysis, simulations, etc. By this
definition, Ada is quite successful and presently in a
state of stability and, yes, even growth.
Of course, the answer also has a dependency on the
definition of "success". If success means getting
coverage (albeit naive) in USA Today, then only Java
is a success. If success means the language is the
first-choice favorite of every hack programmer, then
C/C++ is surely the only success. I prefer to think
that success means satisfied customers, growing sales,
and profit. In this regard, all major Ada suppliers
are reporting stability or growth (though I have no
idea whose products are profitable) and my sense is
that Ada users are much more satisfied with the
products of 1998 than those of 1988.
As far as hard data is concerned, I am not in any
position to speak for any other Ada companies, but
I can tell you some factoids that I reported at
SIGAda last week. Revenues of my product line have
increased 35-40% per year for the past three years
(note: that's before, during, and after the drop
of the 'mandate'); the number of people downloading
ObjectAda is up 225% of the same period last year
(and it was already a lot last year); we now have
in excess of 135,000 units in distribution; and, in
the embedded realm, Ada is identified by Embedded
Systems Programming as one of only three languages
exhibiting growth (the other two are obvious
enough - C++ and Java. C is on the decline).
Ada is steady in terms of percentages, but the
market size is increasing.
Hope this helps.
-- Dave Wood, Aonix
-- Product Manager, ObjectAda for Windows