Re: Investigating Ada

[<adaworks@sbcglobal.net>]

"Jim Rogers" <jimmaureenrogers@att.net> wrote in message
news:6Zm1f.116464$qY1.5654@bgtnsc04-news.ops.worldnet.att.net...
>
> "Phoebe" <no@nospam.net> wrote in news:4cE0f.1602$ir4.1138@edtnps90:
>
> > I'm going to learn it anyway, but I'm just hoping that people could
> > comment on its popularity, and also how it has kept up against the
> > other languages which surely must have incorporated a lot of its
> > advantages by now. Thanks
>
> Interestingly, very few languages have incorporated a lot of Ada's
> advantages by now.
>
Nice list of Ada's benefits in your post.   It seems that most programmers
are more interested in convenience during development than they are the
long-term health of a software product.  The latter concern should be
a management issue, but not many managers are prepared to understand
the problem well enough to choose Ada.

I am currently teaching a graduate seminar titled "Software Evolution."
We are examining the issues related to the lifecycle of a software
system, including how to plan for its continued adaptability to new
user requirements, new environments, and new hardware.  Sometimes
this is thought of as maintenance.

Planning for software evolution, which manifests itself in many forms and
technical demands, is becoming more and more essential.    One approach
is to simply plan to rewrite the code every so often.  This is currently a
popular choice among those who write software in HTML, XML, scripting
languages, etc.   However, rewrite is not a good choice for the majority
of software systems.

As we examine the language alternatives, Ada comes through as a sound
choice.   It is becoming increasing clear that, while extensible software can
be written in most contemporary languages, the very design of those languages
discourages this kind of planned evolution.   Java and Eiffel are fairly good.
C++, as currently practiced and programmed, is horrible.  I suppose C++
is not to blame, but its practitioners seemed determined to prevent anyone
from understanding their code well-enough to extend it.

So, as we study the problem of software evolution, Ada is clearly a good
choice.

Richard Riehle