Paul Everitt <paul@digicool.com> wrote:

Here's an analysis of the OMG's plans for CORBA 3:

http://www.objectwatch.com/issue19.htm

Though obviously subjective and impassioned, I agree with most of what the author says. The OMG has abandoned language-neutrality.

I was sad to see Richard Soley, whom Jim and I worked with 3 years ago, try to tie application servers, Java, and the OMG together. At any rate, CORBA's current level of adaptation is quite low, and now that OMG has become a marketing organization for Sun, I think it's time to dispel any notions that CORBA support is critical.

Just my $0.02.

--Paul

As an active CORBA developer in 3 languages (C++ / Java / Python), and one hard at work on bridging Zope to my CORBA stuff, I have to differ. Sessions' anti-CORBA bias is well-known within CORBA circles: he was lead architect for the Persistent Object CORBA spec, which was roundly rejected by implementers after his team delivered it, late and crippled. He has been actively promoting COM/DCOM ever since, is a frequent presenter at Micros~1 FUDfests, etc. CORBA is not "about" application servers, nor has it abandoned language-neutrality (the C++ mapping is the most widely used, and probabaly will remain so for the next 3-5 years in production systems). We should take pronouncements of ivory tower theorists with a large block of salt; the community is dedicated to the "rough consensus and _working_ code" model of standards development which enabled the Internet protocols total victory over the "sanctioned / blessed" ISO standards. The OMG is hardly a marketing arm for Sun; standards bodies don't work like that. -- ========================================================= Tres Seaver tseaver@palladion.com 713-523-6582 Palladion Software http://www.palladion.com