Re: Ada

[marc@dumbcat.UUCP (Marco S Hyman)]

In article <7088@hubcap.clemson.edu> billwolf%hazel.cs.clemson.edu@hubcap.clemson.edu writes:
       Non-DoD US users include General Electric, 
       Boeing Commercial Airplane, etc.; other companies, such as Arthur 
       Anderson, are considering converting to Ada now that the appropriate 
       compilers and tools are available.  Not because the DoD is using it,
       but as a simple result of business and engineering considerations. 

Large corporations such ad GE, Boeing, Arthur Anderson, etc., are not known
for state-of-the-art, pushing-the-envelope solutions to their problems.
These are the companies that buy IBM.  Ada is safe, approved by the
government, and, as I once heard Larry Rossler of the HP Language Labs
describe it, doomed to success.  Too many companies are throwing money at it
to fail.

       The experience so far has been that "once a team moved to Ada,
       they stayed with Ada.  Once Ada had been successfully used within
       a company, its use proliferated to other projects within the 
       company and to subcontractors and suppliers of that company within
       the industry."  (Ada Letters, Vol. VIII, No. 5, page 15).

I don't have any facts or figures but I suspect from my own past experience
that this is true of every successful tool/method introduced to a large
corporation.  A manager that tries something new usually goes out on a limb
and therefore works twice as hard to make it succeed.  When it does succeed
that manager makes sure all of his bosses know that he was the hero that
finally brought a software project in on time and within budget.  Since
software projects are not usually on time or within budget the new
tool/method is seen as the solution to all software woes.  Alas, it's
usually just better management plus the productivity increase that comes
with doing something new that really saved the day.

I'm not saying the new tool/method, be it Ada, C++, OOP, or whatever, does
not help.  I just wish that it wasn't seen as a panacea for all that ails
the current problems in software.  (BTW:  the current problems in software
haven't changed, IMHO, in the last 20 years -- there's just a lot more
software.) But I've changed the subject.  Sorry.

C++ and Ada are just tools, folks.  Sometimes they are the correct
tool for the job.  Sometimes they are not.

// marc
-- 
// Marco S. Hyman		{ames,pyramid,sun}!pacbell!dumbcat!marc