Re: Ada is getting more popular!

[Georg Bauhaus <rm.dash-bauhaus@futureapps.de>]

On 25.10.10 12:33, Dmitry A. Kazakov wrote:

> Redistribution is the key. You pay/invest in not what you get/sell. It is a
> distorted relationship between the producer and consumer.
> 
> The point is that service is secondary to the product.

I imagine that theorists of how-to-build-software-business will
emphasize differently:
Mathematically, the most important good is the one that
produces the highest profit, respecting the limiting factors
such as law, strategy, or other obligations.  Criteria such
as how-to-find-a-societal-optimum in a market economy
seem to be artifacts of macro economy without positive
manifestation anywhere.  Redistribution is a multivariate
function of n-compiler : m-customers, ceteris paribus.

Can you name the latter?

> It is a parasite living on the host.

(So he who rides a horse is a parasite? :-)

> The success of the former [parasite, service] means nothing good
> for the latter [host, compiler].

The above is a value judgment on compiler business:

Assume that a customer wanting to buy an XYZ compiler
gets a better deal if

- the language is simple, efficient, and consistent.
- the compilers just work.
- both are true for years.

The above customer's desire will, however, fail to influence
the compiler market, or the language.  For one thing, customers
may be competitors themselves.  Consequently, even when a language
is complex, even when compilers need constant attention, and thus
create substantial cost, customers will still not, in effect,
unite to obviate either complexity of language or quality of
compilers.  Who does, then?

Ages old rhetoric will assist either side.  Every forty-something
will have little difficulty explaining that anything of the above
is not really important, too principled, "get a life", family is
what matters, business is corrupt, anyway, the compiler business,
too, needs an income, etc etc.

Does it matters little, then, when some SPARK project seems to shine
even when seen from an economic perspective, being (almost?) under
budget, and in time?


> That this model of software development (not only compiler development) is
> not socially/economically sustainable is obvious when you consider present
> software quality (miserable), the types of software being developed (mostly
> useless/damaging), the amount of resources spend directly/indirectly on
> software (huge waste).


Do you have the numbers of compiler makers who sold
nothing but compilers in, say, the late 1980s versus
the same numbers as of today?  Numbers of licenses sold?