As you may have heard, the next version of SQL Server (code named “Yukon”) has a name. The release is now formally called SQL Server 2005 to match Visual Studio 2005. A new schedule was announced that pushes the beta cycle out as well. Beta 2 will be widely shipped in the summer of 2004. Beta 3 will be shipping in the late 4th quarter and the GA release will ship in the first half of next year.
Most of the users polled in our survey on SQLServerCentral.com are fine with the delay. The general consensus was focused on quality and I heard from most “get it right, not fast”. I couldn’t agree more with this myself. But I do disagree with the strategy that Microsoft has cornered itself into.
SQL Server 2005 will be a dramatic shift from SQL Server 2000. The interface has changed and there are too many new features to list here. Those who know SQL Server today will have to relearn a lot of the application. The sleeper hit is going to be Data Transformation Services (DTS), which has experienced a complete overhaul.
This is too much change for one release and it’s no wonder why the release is being pushed into 2005, making the cycle 5 years between releases. SQL Server users were thrown a bone throughout the cycle with product additions like Notification Services and Reporting Services but no minor releases.
As a SQL Server enterprise, I would much prefer to see smaller more manageable releases that are more like a light meal than going to the dinner buffet. I would like to see a release every three years that focuses on minor touchups and a major overhaul of a few components. The net result would be the same of having all the features by 2005 or 2006 but I wouldn’t have that unpleasant feeling like I do when eating too much at an all you can eat buffet.
In my ideal world, Microsoft would release a SQL Server 2003 which would change the interface to Visual Studio, make some minor tweaks to the relational engine and the major updates to Analysis Services and DTS. SQL Server 2006 could then make the remaining changes that were left out of the first release and finish the overhaul of the relational and storage engine with items like database mirroring and partitioning.
SQL Server needs new features faster than every 5 years. With competitors like MySQL, Oracle and UDB nipping at SQL Servers toes trying to grab the Wintel enterprise, this current 5 year cycle is unsustainable. I’m all for slowing down to focus on quality and security but it’s a lot easier to do quality inspections on a car than an aircraft carrier.
Let us know what you think by clicking the Your Opinion button below.