I really agree with this concern. Corba has a huge learning curve and
some of it's component standards are much better than others (I curse
the C++ bindings daily). While Corba doesn't have to be large and slow,
some of the implementations are.
If you want a self demo and to learn a little from something pretty
concrete, I highly recommend CorbaScript
http://corbaweb.lifl.fr/CorbaScript/
It's an interpretted language that can access and define Corba objects.
The samples that come with the interpreter show how you can "do corba"
with "one line of code".
-Brian
Greg Kreis wrote:
>
> Does anyone know of a concise example of a CORBA based system that is described
> on the net? I'd like to get a look at a couple of designs to get different
> ideas on how folks decided to arrange their model's classes to be served up via
> CORBA. What sorts of CORBA success stories are out there?
>
> I went to a presentation of ObjectSpace's Voyager ORB Tues. evening. They
> talked about how their product was able to bridge/adapt to DCOM, CORBA and EJB.
> They also pointed out, for instance, that to access a CORBA object might take
> about 10 lines of code where they require one.
>
> One thing that struck me about the presentation was how intellectually
> interesting it is to talk about all these layers of technology, yet how it might
> not pan out in the end due to practical matters. Running lots of
> services/servers all linked together over a wide variety of connections (LAN,
> Inernet, etc.) and getting them all to come together in a user's desktop app
> reliably and quickly is a daunting task.
>
> Perhaps we are at the stage of distributed systems where we were with LANs in
> the 1980s. Lots of promise, but lots to do before everyone naturally assumes it
> is there and working.
>
> I hope this does not take us too far off the main theme here, but I personally
> don't have work experience in using or running distributed object systems and so
> I don't know the reality that is hidden behind the media's articles.
>
