Re: MicroSoft to purchase Aonix?

[jerry@jvdsys.nextjk.stuyts.nl (Jerry van Dijk)]

In article <5mkgrl$akp@ion1.ionet.net> mbates@ionet.net writes:

I know the feeling... <sigh>

>"Any improvements in new releases of NT will be immediately accessible
>in C++, but we might have to wait a year or so before Aonix has
>upgraded the Ada compiler to reflect the changes."

Ok, what improvements are we talking about that would require an upgrade
of the compiler ? Instead of adding/changing a spec file and an import
library ?

Does release evidence show that new features immediately show up in VC++,
the next month or so, or is there some lagging behind too ?

Are they sure that all the improvements they obviously foresee in NT are
actually needed/used ?

>"First it was Alsys, then Thomson, then Aonix.  Seems pretty unstable.
>I know MicroSoft will be around in 10 years.  How do I know this
>little Ada company will still be around?"

Well, first of all, Aonix is not the only game in town when it comes to
NT Ada compilers, there is at least GNAT to consider.

Portability between Ada compilers can be expected to be _a lot_ better
then between versions of the same C++ compiler...

How do they know MS might still be around 10 years from now ? Let alone
NT. And will this NT still resemble the current version ?

And for that matter, what is C++ going to look like in 10 years time ?
Will VC++ still be able to compile currently written software...

Look back 10 years, was this Ada compiler available back then ? (hint:
yes) and is the software written back then stil compile-able today ?.

>"Countless millions of dollars are being spent to develop tools, class
>libraries, etc., for use in NT development.  We won't be able to use
>them if we go with Ada.  The Ada market is too small to bother with
>for software development tool makers."

Why do they think that a lot of NT tools are not going to work with an
Ada system. Or do they mean C++ tools ?

If the latter, who is telling them a) you cannot use them from Ada, b)
that you will need them in Ada too, c) that there will not be specific
Ada NT tools too, d) that all these tools will be relevant.

>"Ada is like Betamax.  It may be technically superior, but it's on its
>way out.  Everyone is doing C++, and we'll be passed by if we go with
>Ada.  Ada is a dead-end."

1. Ada is not like Betamax since it is currently available on almost all
   relevant hardware platforms;

2. Yes, Ada is technically superior.

3. Where is the evidence that 'Ada is on its way out' ? All _facts_ I am
   aware of show the contrary.

4. Everyone is not doing C++. Everyone is doing Java (better use this
   then), Powerbuilder, Cobol, Excel, etc.

5. How are you going to be passed by using Ada ? In what sense ? Will
   C++ develop further into a better language than Ada, are they willing
   to spend the amount of money needed to keep abreast of all those
   changes ?

6. Business-wise Ada's advantages reflect on the bottom line, going with
   VC++ means spending a lot of money to stay current with whatever MS
   decides to do next.

>"Some of our outside customers will be maintaining the simulator
>software themselves.  They've heard how wonderful C++ is.  They won't
>want to buy a simulator programmed in Ada."

If these customers decided against a better produkt written in Ada
because they 'heard how wonderful C++ is', please send the names and
addresses, since I know a few other nice deals for them too...

>"We don't want to swim against the stream."

Right, and the company that builds its produkts just like all its
competitors do is the one that is going to capture the market...

>"Everyone's learning C++; no one's learning Ada.  We'll always have an
>easier time hiring C++ programmers than Ada programmers -- and they'll
>be cheaper too."

If no one is learning Ada95, then how do they explain the rapidly rising
number of books teaching Ada ?

Yes, there are more C++ programmers then Ada programmers. But are all
C++ programmers created equal ? (ever seen an experienced gpp programmer
wrestle with Visual Age ? <g>).

And are C++ programmers really cheaper then Ada programmers ? How did
they measure this ? Just salary ? Is that being business-wise or
foolishly believing that all other factors are equal ? For example,
might measuring productivity no show that over time Ada programmers are
actually cheaper than C++ programmers ? (hint: rational study)

>"C++ is the native language of the NT platform.  You have to write
>device drivers in C++."

Nope, the native language of NT is C, so are its device drivers. But
even if this were true, what relevance does this have for application
development ? Especially since Ada is interoperable with C (and GNAT:
C++), and the NT platform has dll's and such.

>"MicroSoft uses C++ for their development.  Don't you think they'd use
>Ada instead if it was really that much better?"

Do they really think NT and Win95 were written in 32-bit C++ ?. Sounds
like MS marketing is even better than I thought.

And remember that MS is going for the mass market, they have no concern
for quality, let alone that their customers will be maintaining MS
produkts.

A valid argument, if MS were selling simulators too, otherwise I do not
see the point.

>Anyone have any good __evidence__ that counters these arguments?  (And
>yes, I've been to www.adahome.com, but a lot of the info on commercial
>use of Ada appears to be pretty old.)

Yes, that's the result of the transition from '83 to '95. It is
being worked on as we speak :-).

Also note that, at least here in Europe, a number of companies that
are using Ada to their advantage don't want this fact to go public,
for fear of waking up the compition.

Good luck!

--

-- Jerry van Dijk       | Leiden, Holland
-- Business Consultant  | Team Ada
-- Ordina Finance       | jdijk@acm.org