🚀 go-pugleaf

Peter van der Linden wrote:

> Look, if the argument has changed from:
>
>     (1) CORBA is better; you should use it in preference to RMI
>
> to
>
>     (2) Everyone should get behind CORBA because otherwise CORBA will
>         lose against DCOM
>
> then at least say so explicitly.   Nobody in the RMI camp is buying argument (1)
> if the real agenda is (2) then let's phrase it that way.
>

Peter, the real agenda is (2).  In our book, we tried to make RMI
programmers aware of the Netscape/Visigenic Caffeine alternative.  We
believe in (1), but I don't want to republish the book here:-)  So (2)
is where this discussion should be heading.  Some of my students are
mortaging their homes, in order to create distributed Java products
(based on Enterpise Beans).  I don't want to see them go bankrupt. :-).
The memory of OpenDoc is still very fresh (may it *not* rest in peace).
I for one learned the hard way never to underestimate Microsoft.  Right
now ActiveXs with DCOM and Viper are starting to look very attractive.
We must work on a credible alternative.  These battles of the ORBs (RMI
vs CORBA) are ridiculous.  They're  based on internal Sun politics.
JavaSoft must pick one ORB and go with it (we warned them about them
this last September).  We shouldn't be made to suffer because they can't
get their house in order. I can assure you that they do not have the
resources to support two ORBs in the style required to beat DCOM.  The
discussion should be: Why not RMI on top of IIOP?

Bob Orfali