From: Jeff Garland (jeff_at_[hidden])
Date: 2005-03-06 22:00:39
On Sun, 06 Mar 2005 18:24:15 -0800, Steven E. Harris wrote
> "Jeff Garland" <jeff_at_[hidden]> writes:
>
> > Also, the CORBA binding to C++ is a pain in my view -- someone could
> > redo it with modern C++ and things would be much better,
>
> Have you looked at ZeroC's Ice?¹
Yes, I read the docs a few months ago and it looks really good. These guys
are incredibly smart about distributed computing and have been deeply involved
in CORBA -- so they know the good, the bad, and the ugly of CORBA. Only thing
is, they want to make money off their knowledge (see the license -- GPL for
open source use, pay them for commercial) -- can't blame em. So this stuff
won't be coming to the C++ standard or any other standard in the near future...
> > but I don't see anyone talking about that.
>
> Doug Schmidt and Steve Vinoski have written several articles about
> the difficulty involved in updating the OMG C++ mapping. As I recall,
> they argued that it's not that everyone doesn't there's plenty of
> room for improvement, but rather that the interested parties had trouble
> finding consensus as to how to update the mapping.
That's too bad because leaving it alone means lots of heartache when doing
CORBA programming -- makes me avoid it when I can. If the C++ mapping were
cleaner I might have a better attitude toward using CORBA because it has some
really nice and essential features for distributed programming.
Jeff