Michael Feldman wrote:
>
> Surf over to http://www.datamation.com/PlugIn/newissue/DATAM.html
>
> and have a look at her article and my sidebar. Write to her if you
> disagree with her outlook on Ada.
>

Here's what I wrote to her:

-------------------------------------------------------------------
Re: Ada and legacy systems

Karen Watterson:

The analysis which concludes that Ada's prospects are
"not good" (5/98) is flawed.  The reason given for this
is that the DoD has dropped the Ada mandate, and as a
result, the DoD's millions of lines of Ada code will
dwindle.

If this is so, why did so many DoD systems avoid the
use of Ada even when it was a DoD requirement?  Why are
many DoD systems choosing Ada 95 for new systems after
the mandate?  And why have there always been (and continue
to be) a significant number of high-profile systems
outside the DoD which choose Ada?

The answer is simple: the mandate is not really the reason
that Ada was successful.  The mandate was always a
"contentious" policy (to use the words of General Emmett
Paige, who signed the memo rescinding the mandate); in fact,
it became something of an embarrasment to the DoD because
it was so widely ignored.  The organizations that were
determined never to use Ada didn't (or eventually didn't);
the ones that found that it met their requirements chose it
consistently.  So do not expect Ada's fortunes to be
radically affected by the lack of a mandate.

I must disagree with your perspective on Ada's prospects.
For real-time systems there is currently no language that
meets the same high standards as Ada 95, and this continues
to be the reason is it chosen for some of the most difficult
and vital systems in the US and Europe.  The reason is
ultimately financial: for those organizations, Ada has
assisted in saving a tremendous amount of money.  Over a year
has passed since the rescission of the mandate, and the Ada
compiler vendors I am familiar with continue to report
strong sales.  Also consider that one of the faster-growing
markets (in terms of percentage of revenue) in the computer
industry is real-time systems.

As one of its designers recently stated, "In some sense
Ada is a "quiet" language. The applications built with
Ada are not in general the widely known desktop or
entertainment applications, but rather the mission-
critical, real-time applications supporting the
transportation and aerospace industries."  (See
http://www.objectmagazine.com/features/9804/taftinterview.html)

Thank you,

Stanley Allen
mailto:[log in to unmask]
---------------------------------------------------------

P.S. to teamers: It stood out at my reading that the
only example she gave in the article of an organization
abandoning Ada was a wall-street firm (I'm guessing
Reuters) which of course was under no DoD mandate to
use Ada, and that the reason given for converting to C++
was that "we can't find qualified programmers" (not
"the Ada tool market is drying up" or "we believe that
using Ada is betting on a losing horse").  About two
years ago you always heard that complaint about
Ada programmers being hard to find; ever since then
it has become obvious that programmers in general
are hard to find, so I believe that reason for
dropping Ada carries less weight.

If the company mentioned in the article was in fact
Reuters, it would be interesting to know if the
conversion to C++ really worked or was comprehesive,
since we've seen at least one post on comp.lang.ada
in the last 4 months from a person at Reuters asking
about how to find some good Ada compilers.

Intriguing, n'est-ce pas?

Stanley