Re: Where's Ada95 when OO languages are discussed?

[Chris Morgan <mihalis@ix.netcom.com>]

David Starner <dstarner98@aasaa.ofe.org> writes:

>> > And some things you say are still wrong.
> You seem to just be arguing here.

No I'm responding to your posts which contained inaccuracies.

> 
> > That's true. However Ada isn't even as visible as it might be given
> > the market value of Ada projects, since a lot of them are embedded
> > and/or defence projects (Boeing 777, USAF F-22, Channel Tunnel
> > signals, London Jubilee line, UKRN Trident sub CCCI - notable but not
> > "newsworthy").
> 
> That's sort of what I meant by mainstream*. The only major program that
> might cross my desktop that's written in Ada is GNAT, to the best of my
> knowledge.
> 
> * In retrospect a very poor choice of word.

Does software that gets you from airport to airport count as
mainstream? Are you limiting it to end-user apps? 

> > Ada is getting on for 20 years, so it has _enough_ history in my
> > mind.
> Interesting. Unfortunately, I don't have your mind on hand, so I
> couldn't consult it. 

Your sarcasm is no doubt amusing, however you started off by comparing
Smalltalk to Ada and saying "Smalltalk has history". What exactly did
you mean by this?

> 
> > True, many OO features are bolted onto the core language in
> > Ada95, but other features have been there since the beginning. It's
> > not "pure" whatever that means, but then neither are C++ and Java.
> 
> "Pure" in OO terms refers to whether it forces everything to be OO. Java
> is generally considered a pure OO language. C++ is not

Yes, Java is more "pure" than Ada95, but it still has non-class
integers, floats etc so it is not as pure as Smalltalk. The reason I
used quotes is that the word "pure" is often used with the connotation
of "better" when OO advocates discuss the relative merits of different
programming languages but that's actually just a matter of opinion. In
actual fact there is still a large body of opinion that a langage that
allows OO styles and also programming "to the metal" is necessary for
many tasks.

> Jargon File:
> Hackers are nearly unanimous in observing that, technically, it [Ada] is
> precisely what one might expect given that kind of endorsement by fiat;
> designed by committee, crockish, difficult to use, and overall a
> disastrous, multi-billion-dollar boondoggle ...
> 
> The canonical reference. Enough said?

Yes, you're a clown. Ok, I'll stop the sarcasm unilaterally now, but
consider this, as amusing as the Jargon file is, are you stating this
is your canonical reference for knowledge on programming languages?

> 
> > Common knowledge is of no interest when the facts are known, and I
> > like to correct such inaccuracies wherever possible.
> It was an explanation about why Ada wasn't considered when other OO
> languages are. "Common knowledge" is very much of interest when
> discussing the actions of "the common man".

There is an effort to spread news of the current, factual state of the
Ada language to the common man underway right now. I don't know how
it's succeeding, but I suppose the people concerned will really have
their work cut out if everyone is going to react like you - "Everyone
knows it's a bad language, why even the jargon file says so". However
I think your description of the common man really only applies to
you. This is just your opinion of Ada based on your own limited
knowledge with the invaluable jargon file to back you up. You do
realise it's not a serious document? I believe there is a document
atacking all progress since Fortran as irrelevant which is also
canonical as net-fluff humour goes.

> > Interesting if relevant, however it isn't.
> Relevant? It was an aside, and marked as such. If you aren't interested,
> just delete it.

It was an aside based on the premise that Ada95 was designed by
committee. Since that premise is invalid your aside is irrelevant in
any case, whether I am interested or not.

Let's move on. I don't want any kind of slanging match, and in fact I
work in C, C++ and Java for my living. In addition to the kind of FUD
you posted, another reason Ada doesn't appear in OO journals is that
OO is still quite small in the markets that Ada was designed for,
whereas it is huge in C++'s market. Dynamic dispatching is rightfully
regarded with some suspicion in a world where a lot of software still
makes do without dynamic memory allocation - it must be justified
carefully. In many ways Ada is an advanced and demanding language and
it was only recently that a variety of fast inexpensive Ada compilers
became widely available - coinciding with the rise of the killer
micros (PCs). In the same space of time it's become clear that Eiffel
has some deep flaws (what is the numerical model of an Eiffel
program?) and that Java will remain uncompetitive with compiled code
speedwise, basically, forever. I probably don't need to mention the
mess C++ is in, so I see the future of Ada as still bright.

Chris
-- 
Chris Morgan <mihalis at ix.netcom.com                http://mihalis.net
     "I don't have time to read a 1200 page book. 
      I am afraid to even let one in my house." 
                                     - Philip Greenspun