Cost of Ada (was - Re: C++ vs. Ada -- Is Ada loosing?)

Cost of Ada (was - Re: C++ vs. Ada -- Is Ada loosing?)

[fred j mccall 575-3539]

In <1992Dec30.040140.10412@seas.gwu.edu> mfeldman@seas.gwu.edu (Michael Feldman
) writes:

>>Basically, the high cost is for the RTE and the cross-
>>compiling utilities, not for the compiler itself.
>>
>This seems a weak justification. I'll agree that Ada may need a more
>sophisticated RTE than C, but is it 25 times more so?

>And the argument about a small customer base is self-fulfilling, as we
>have discussed many times in this group. 

And I will assert that the reason the costs for such Ada systems don't
come down is, in fact, the Mandate.

Consider.  Most of the current customers for such systems are doing
Defense work, where Ada is mandated.  They are, in effect, consumers
in a captive market.  This artificially pushes the demand curve to a
position where demand is artificially high at each market price and
the elasticity of demand is quite low.  Hence, the supplier in such a
market will exercise what is, after all, only good business sense, and
will price their product artificially high in that market.
Arbitrarily lowering their price in that kind of captive market
situation leads to lower profits (since they make less on each of the
'mandated' purchases).  This is why you will likely never see anything
like a 'Turbo Ada' while the Mandate is in force, unless the
government funds the thing itself (trying to correct market
distortions caused by government intervention through the use of more
intervention). 

In this kind of market, at least some people who would otherwise
choose Ada (if the software were available at the real free market
price rather than the artificially high price created by captive
demand from the Mandate) find themselves priced out of that decision.
While the increase if 'free choice' users might more than make up for
the loss in profits from 'mandated' users if the price were to be
lowered, in the short run this will be a losing strategy; so it isn't
going to happen, more than likely.  Thus, the Mandate itself helps
create the situation where only people who are forced to use Ada will
do so (because of the price), which creates the necessity for the
Mandate (in somebody's eyes).

Solution?  Ditch the Mandate.  Let Ada compete on the basis of its
merits in the same kind of marketplace that other languages do.  It
may win and it may lose, but I think one thing you'd see is the prices
of Ada systems coming down to compete with other languages and
language systems that provide similar functionality.  There would
perhaps be an initial drop in systems being built in Ada as people
previously subject to the Mandate selected other languages, but I
suspect that that might be more than made up by other systems being
built in Ada as the prices for Ada language systems coming down under
market forces and people decided to use it.

-- 
"Insisting on perfect safety is for people who don't have the balls to live
 in the real world."   -- Mary Shafer, NASA Ames Dryden
------------------------------------------------------------------------------
Fred.McCall@dseg.ti.com - I don't speak for others and they don't speak for me.

Re: Cost of Ada (was - Re: C++ vs. Ada -- Is Ada loosing?)

[Gregory Aharonian]

>Consider.  Most of the current customers for such systems are doing
>Defense work, where Ada is mandated.  They are, in effect, consumers
>in a captive market.  This artificially pushes the demand curve to a
>position where demand is artificially high at each market price and
>the elasticity of demand is quite low.  Hence, the supplier in such a
>market will exercise what is, after all, only good business sense, and
>will price their product artificially high in that market.
>Arbitrarily lowering their price in that kind of captive market
>situation leads to lower profits (since they make less on each of the
>'mandated' purchases).  This is why you will likely never see anything
>like a 'Turbo Ada' while the Mandate is in force, unless the
>government funds the thing itself (trying to correct market
>distortions caused by government intervention through the use of more
>intervention). 

This suggestion and others dealing with DoD Ada software policies pose very
interesting microeconomic conditions that have been totally ignored by the
DoD over the past ten years.  Many free market concepts - competition,
marketing, supply and demand, cost curves, etc - are for the most part no
reflected anywhere in DoD software policy planning.  There are no accounting
data sets to assess such questions, no microeconomic models to apply such
data to, and no interest inside the DoD, in particular the STARS effort,
to care or be concerned about such questions.  At least the Russian central
planners had economic models of what they were trying to do, even if the
models were lousy economics.

The problem with the Mandate is that it is part of the big Lie: that anyone
can claim anything about DoD software policies, knowing that no one downstream
will be able to veryify or disprove such claims, or remember what the original
claims were.  The STARS program, with its endless reorganizations and new
sets of claims, is a classic example.  Ada is just a side issue to the DoD
brass (my face is still blue holding my breath waiting for Strassman to
publicly mention Ada in one of his interviews).

Greg Aharonian
Source Translation & Optimization
-- 
**************************************************************************
Greg Aharonian
Source Translation & Optimiztion
P.O. Box 404, Belmont, MA 02178