Thursday, May 27, 2004

Microsoft Kills Kodiak Exchange Server

Microsoft Kills Kodiak Exchange Server "In a surprise move, Microsoft revealed yesterday that the company plans to kill the next major version of Microsoft Exchange Server, code-named Kodiak, and will instead plot a future course of smaller upgrades to Exchange Server 2003. These upgrades will, over time, comprise what would have made up the feature set for Kodiak, which Microsoft originally planned to release in 2005.
"We're actually going to stop using the Kodiak name," Microsoft Corporate Vice President David Thompson said. "But there is a set of things that we're still working on and that we will announce in steps. The next major technology, Exchange Server Edge Services, will come next year." Another Kodiak technology, the Microsoft SQL Server-based data store, will probably be postponed until the next major Exchange Server version, now due in the 2006 to 2007 time frame."

That's a misleading headline -- Microsoft killed the code-name, since it had managed to totally confuse the market about plans for the next major release of Exchange, but it hasn't changed the overall product strategy.


Anonymous said...

If it were true that they hadn't changed overall product strategy, why the dodge over the plans for SQL, timetable for a next major release, etc.?

pbokelly said...

Sorry -- I should have been clearer; Microsoft has of course changed the Exchange strategy in significant ways since Exchange 2000 was announced, primarily in dropping the collaboration-related parts of Exchange (replacing them with Live Communications Server, Live Meeting, etc.).

But the key thing about Kodiak was the plan to move to SQL Server, and that's still the plan; it's just taking Microsoft a lot longer to get there than they originally planned. To be fair, they're also doing a lot of useful interim stuff, e.g., the edge server model (which, with managed code APIs and SQL Server as the queue storage engine, is rather Kodiak-ish in spirit). In any case, I think they were right to shoot the code-name, and to reset the plan rather than sticking with arbitrary historical dates that were made in a very different set of market circumstances.

In fewer words: "mistakes were made..." but it's better to press the reset button and fix the plan/schedule than to stick with a plan that wasn't going to work as hoped.