Re: Ada image problems, was JGnat support.

["Joachim Schröer" <joachim.schroeer@dornier.eads.net>]

Ingo Marks wrote:

> Joachim Schr�er <joachim.schroer@dornier.eads.net> wrote:
> 
> 
>>The current public version does only support j2sdk1.2 (currently we have 
>>1.4.0). From ACT-Europe I got the information that JGnat will not be 
>>maintained in future due to lack of customer interest.
>>We were interested in Ada -> Java technology that has future. Our current 
>>project demands Java and we have a few hundred thousend loc of (not very 
>>nice) Ada83. I would prefer a transition to Ada95 and a parallel 
>>improvement of the code but we need an Ada95 to bytecode compiler.
>>Has anyone information about maintenance of other Ada->Java products like 
>>Aonix ObjectAda/Averstars appletmagic or similar?
>>
>>J. Schr�er 
>>
> 
> If ACT-Europe is not willing to support JGNAT anymore, they would possibly be 
> willing to put JGNAT into Open Source. Would that make sense or doesn't it 
> matter due to lack of commercial support? I think it's better to support 
> JGNAT by the public than to let JGNAT disappear into oblivion. Another choice 
> would be to write a Ada 95 -> Java compiler (and/or Ada95 -> C#) but that 
> would be pretty much work ;-)


I don't think Ada will disappear. There are too many large systems out 
there that need maintenance. And there are domains where Ada also will 
be selected for new projects.

But there are more and more projects, not safety critical, no realtime, 
where new stuff is written in Java or C++. The decision makers are told 
since years that Ada will die and they assure that this is a self 
fullfilling prophecy.

The young colleagues that come from university only know C++ and Java. 
Some have difficulties to read type declarations (Is Pascal not taught 
nowadays). If you tell them of Ada the usual question is: "Is it still 
used?". There is absolutely no interest in learning, reading about Ada. 
It's like me in school, I hated learning latin cause noone speaks it.

I put the Ada/Java comparision papers into our intranet but noone took 
the hour reading one of them.

In my environment (and I've seen a lot of projects) I see a major 
problem with Ada. The so called Ada experts are often out of date. They 
only know Ada83. They often have no clue of OO, protected types, 
bindings and all this new stuff. They don't read newsgroups, magazins 
etc. On the other hand you have these young Java guys. They are up to date.

And I've seen so many huge and absolutely horribly coded Ada(83) 
systems. The only reasonable conclusion: The authors of that code knew C 
or Fortran, where forced to write in Ada and did it without any ambition 
and without reading a book.

I've no idea how this could be changed. I'm tired of fighting these pro 
Ada wars.


> 
> I am sorry about that most programmers prefer light-weight languages like 
> Java and C# instead of Ada 95. Of course, they seem to be more comfortable 
> and easier to learn. But my experience with Ada is far better that with all 
> my previous languages Perl, PHP, Java, C++ etc. The larger the projects the 
> better the life with Ada. Now the IMHO very best language seems to pass away 
> - slowly but surely. It's a shame! But there is some hope: In last weeks I 
> discovered some new websites regarding Ada 95 (for example, 
> http://www.ada-deutschland.de in German).


This site surely will not help Ada very much!


> 
> By the way: Does anyone know about any Ada support for the Microsoft .NET 
> platform?
> 
> Ingo
>