Despite an announcement, 'Kilimanjaro' may not be the next SQL Server

There's a difference between building a new database engine, and building tools that are bundled with an existing database engine for a new product. So even analysts may be surprised to learn the next SQL Server is not two years away.

At a conference in Seattle yesterday that was apparently accompanied by at least one demo that was not on its original schedule, Microsoft made mention of a product with the code-name "Kilimanjaro," in association with SQL Server.

Since "Katmai" was the code-name for the last release that just made it out the door -- albeit six months later than hoped -- there was immediate assumption that Kilimanjaro (a taller mountain) would refer to the next version of SQL Server. It didn't help that a Microsoft press release referred to two new business intelligence features being developed for SQL Server, using the following phrase: "Upcoming in the next release of Microsoft SQL Server focused on BI -- code-named 'Kilimanjaro."'

The phrase is accurate, if you understand that the code-name refers to the BI features, not the database.

Within minutes, both bloggers and reporters seized upon the phrase as though the company had actually offered a peek at some sort of "SQL Server 2010," and some had actually reported that a database engine for two years down the road had actually been demoed.

The confusion rose to such a level that some of Microsoft's own developers associated with the real Kilimanjaro project actually found themselves correcting their company's own public relations. Andrew Fryer, an evangelist with Microsoft UK's Developer & Platform Group, wrote this afternoon, "Kilimanjaro is a bigger mountain than Katmai, but the project is not the next release of SQL Server 2008, it's a BI set of features that integrate into SQL Server and comprises two projects that you may have heard about."

Those two projects include one called Project "Madison" (yes, another code-name), which is the culmination of a technology that Microsoft acquired just last July, in the buyout of a firm called DATAllegro. It's a data warehouse (DW) technology that compresses the tools needed to make major queries of huge repositories of data. Attendees at Monday's show report having seen a demo of Madison which successfully ran a query that returned one trillion (that's with a "t") records from a 100 TB database, using ordinary PCs in mere seconds.

The other is a concept called "Gemini" which is a set of analytics and reporting tools that will enable individuals to get a more insightful read about the contents of large, dynamic databases. Microsoft customer advisory team member Richard Tkachuk attended a demo of Gemini presented by Principal Program Manager Donald Farmer.

In his blog yesterday, Tkachuk wrote that Farmer "started with 20 million rows in memory on a no-frills PC and built a model from scratch. Even with 20 million rows, interactive ordering, filtering and windowing and pivoting was instantaneous. It's difficult to compare how much simpler it is building a model with actual data than it is building one with abstractions and not seeing the result at the end. You might remember the days before WYSIWYG when documents were built with formatting and font codes and not seeing how it would look until it hit the printer. This is doing the same thing to data analysis with non-trivial amounts of data -- WYSIWA (what-you-see-is-what-you-analyze)...And -- this bears repeating -- it's all in Excel."

It does indeed bear repeating: It's an Excel add-on. It's due to be packaged along with a business intelligence version of SQL Server 2008, currently slated for 2010.

So why does the distinction matter? If there were a new version of SQL Server slated for as soon as 16 months away, the beta cycle for it would need to start...about now.

"Like a lot of conference announcements none of this is going to be ready for some time and even the earliest betas are a long way away," wrote Microsoft UK's Andrew Fryer, "so the only impact it should have on deployments today is that this will just add more useful stuff to the BI projects built on SQL Server."

Among the first Internet sources to correct the mistaken report were independent blogs whose interest in the non-existent next beta may have led them directly to the bad news. In one case, the correction was accompanied by a :-( sad-face emoticon.