Good Move Mr. Litwack. If the Novell BOD had any balls they would have kicked the MESS-MAN long time ago and appointed Chris Stone as CEO. Messman is ruining the spirit of Novell.
The stock is at 6 instead of 20 due to the MESSMAN non-event. This CEO is a fake, absolutely useless. He has been on the Novell BOD since 1986 ( I believe) . When Novell took over Cambridge TP, Messman was asked why Cambridge didn't use any Novell products he replied: "I didn't know they have such great products ( after 15 years on the Novell board). Gimme a break. This man has to go !

Good for Mr Litwack - he always was a decent guy. Maybe Stone will have the last laugh after in all in this tawdry corporate tale.

I can not agree more with the advice at the end of the previous comment by Vance Lister : "Whether an institution implements monolithic, client/server, middleware or SOA, make sure the data format is consistent; reusing a data definition will be more cost effective than reusing code."
High level of reuse of data definitions can only be achieved if common data dictionaries are used to model and implement databases, middleware, services and user interfaces. The product I am working on - MetaBoss offers facilities to maintain enterprise data dictionaries and reuse data definitions across all systems and layers.

We look to SOA because Middleware has failed us. An enterprise has hundreds of computers on desks and in departments, running vastly disparate applications from many vendors. Middleware exists to get data from one place to another and ultimately gather data for integration. Unfortunately, the cost of middleware, interoperability and integration now costs more than the applications themselves (certainly in the financial world where I work). SOA holds out the hope of reducing the number of applications by providing a single service, enterprise-wide, where duplicate applications previously existed. All SOA enabling environments offer the opportunity to mix and match services from different vendors for use throughout the enterprise. Unfortunately, when vendors offer fifty web services instead of one application module, the cost of web service integration will be increased even more and institutions will eventually declare that SOA is dead and some other technical integration platform will be the new silver bullet. My advice, whether an institution implements monolithic, client/server, middleware or SOA, make sure the data format is consistent; reusing a data definition will be more cost effective than reusing code.

"People buy cars" - is probably referring to people who need cars. People who need middleware will buy middleware. People who need SOA product will buy SOA product and will probably get some sort of middleware within it. So it looks to me that SOA product vendor will need to buy middleware. We have had SOA for years. CORBA and / or COM/DCOM/ActiveX could have been used to implement SOA within organisation for long time. But never before we have heard that middleware is dead. So what is so different about SOAP that now gives vendors right to say that middleware is dead ? Will they not use a database ? Will they not have distributed architectures inside to implement their Services ? Will they not have transactions ? I think they are just trying to "segment" the market and carve up a niche for themselves.

Great question: we asked IBM in our May issue, yes. Here's the relevant exchange, with John Swainson the GM in charge of the entire WebSphere product set...

JDJ: [Sun's] Jonathan Schwartz went on record in JDJ as saying "middleware is history." Clearly that wasn't meant literally, but he was saying that end-to-end "systems" will supplant it as a focus. Is the IBM view that middleware, on the contrary, is just beginning?

Swainson: Saying that middleware is "history" is laughable. IBM has tens of thousands of customers who need and use middleware for transactions, data management, development tools, systems management, security, and collaboration in a heterogeneous systems environment. The WebSphere platform experienced 12% revenue growth in 2003 over the previous year. The WebSphere platform grew 24% in 1Q04, marking its twenty-second consecutive quarter of revenue growth.

Plus we have already this year carried extensive interviews with BEA, and with David Skok of Matrix Partners, who invested $10M in JBoss.

Um, is this the new marching order for the Java elites? "Go out there and tell everyone that middleware is dead so we can push this instead..."
First Jonathan Schwartz writes an article in the JDJ, proclaiming the end of middleware (in which he completely failed to tell us why), and now this guy from Novell.

Whatever. Have you guys asked what JBoss, BEA or IBM has to say about this?