It has been almost six years since Microsoft released a new version of SQL Server, and now after so much discussion it appears we finally have a new one.
They should have called it SQL Server Barely 2005. In fact this reminds me of the automobile companies that start selling next years designated models this year. However in that industry we usually get some good deals on the prior years model; in the software business all we usually get is a painful upgrade along with the inevitable de-support notice.
So is SQL Server 2005 a significant release? It is for Microsoft but for the average company? The one significant fact about this release I think is how little the market cared that it took so long.
In an industry that is constantly pushing new and improved software at you, Microsoft defied conventions. Oracle and IBM have each had four significant releases including two major version upgrades since SQL Server 2000 was released.
To be honest, only the diehards care about new releases. They cant wait to get their hands on the new code and try new features. The average company views a new version with about the same enthusiasm that I have for raking leaves. It simply means that the clock is ticking to an inevitable point in time where action must be taken, not because we care to take that action but because some outside force, such as nature or software companies make us react according to their timetable not ours.
In what some may view as a perversion of natural law, SQL Server has actually benefited from its elongated release cycle. The market has grown much more comfortable with SQL Servers stability and its capabilities and because of this, Microsoft has been able to grow its market share. So in a time when most vendors use new releases to spur additional sales, Microsoft spurred new sales by being a contrarian. Now I doubt this was the original plan. After all, based on typical software development cycles, its probable that planning for this version began in 1998. Has Microsoft changed so much internally?
In a 1998 expose of sorts, a book entitled Barbarians Led By Bill Gates tells of an amusing clash of cultures as Microsoft was hired by IBM to create the GUI for OS/2. As I recall it, the author, a Microsoft developer, talked about how obsessed the IBM team was with releasing perfect code and how Microsoft developers were obsessed with speed to market. Apparently that need for speed has disappeared from the Microsoft culture. About the only thing that comes quickly from Redmond these days are press releases trumpeting some new code-named piece of software.
For customers who have paid for Microsofts software assurance program, six years between releases should be an area of concern. These programs often cost organizations 20 to 25 percent of the licensing costs annually. Sure, they help companies predict future costs through a subscription-like service, but at those rates a six-year wait mitigates the advantages of such a program.
While concern over the huge time gap between releases has been almost nonexistent, SQL Server users should at least be aware of the implications. Why did it take so long? Well that question I imagine has many answers, but users should not ignore Microsofts moves in other areas. Microsoft is assembling an integrated software stack, and SQL Server promises to be a central component, deeply embedded in other products and services such as e-mail, directory services, collaboration and development tools.
The question, of course, is will the whole be superior to the sum of its parts? While tight integration with other Microsoft products will undoubtedly bring welcomed functionality and interoperability, what will it mean in terms of future releases and upgrades? Is Microsoft pushing us towards a future where integration is so tight that all pieces of the stack must be upgraded at the same time? For those of you who fill infrastructure administration roles you can already envision the full employment consequences of such a future.
SQL Server 2005 does have some cool features. Noteworthy ones include database mirroring, concurrency and recovery improvements, scalability and availability improvements, security and BI features. Its a big release and one that will take users several years to digest.
I do not expect major adoption to really gain momentum until 2H07. This is due in some cases to the uncertain nature of SS2005s ever changing release date. Most companies I have spoken with were reluctant to put aside budget money for the upgrade because they were not sure when or if it would be released. Secondly, most companies shy away from newly released software from anyone. Now that Microsoft has made inroads into major data center applications, I expect they might be much more cautious as Microsoft still has a lingering reputation for poor quality first releases.
I for one always recommend caution especially when there is no pressing need to move quickly. So lets raise a toast to SQL Server 2005, our long wait is finally over. Of course if you never cared anyway its fun to have a reason to toast just the same.