Re: Is JGNAT production ready?...

["Rob Veenker" <veenker@xs4all.nl>]

I have used JGNAT (supported version) on a project that involved running an
application on a handheld PC using a JVM. There was a mix of Java and Ada
source and after some puzzeling at first, I managed to get Ada to call Java
(including calls from tasks) and Java to call Ada packages (classes).
It saved me the trouble of learning a new language and I too really needed
the tasking support from Ada since I was implementing a proprietary protocol
for driving a communications radio. The main penalty of using JGNAT is that
you get quite a few runtime Java classes to link with your code. This was of
some concern to me because a PDA doesn't have much :-)
The project was relatively small but I got a good impression of JGNAT. It
matched up to my expectations.
I haven't had a look at the public version so I can't comment on that one.

Regards,

Rob Veenker

Andrzej Lewandowski <lewandoREMOVE@attglobal.net> wrote in message
news:3a53bed0_3@news3.prserv.net...
> I am working on some Java application that is heavily using threads.
> Programming multithreaded applications in Java is, ehem... sorry...
> no fun. Therefore, I have undusted my Ada83 skills, upgraded to 95
> and I am thinking about using JGNAT for this part of application that
> uses threads, with the rest in plain Java.
>
> This is commercial application, and I have no room for mistakes.
> Mistakes would be expensive. Therefore, the question(s): what is the
> quality of JGNAT? Is here anybody who has experience with using
> JGNAT for medium-size real life projects? Linking JGNAT modules
> with Java modules? Quality of JGNAT threads?...
>
> Other question: is GNU JGNAT as good/bad as the commercial version
> of JGNAT? Are they the same JGNATs?...
>
> Andrzej Lewandowski
>
>