Picture a barren landscape. Two forlorn and confused SQL Server DBAs wait. And wait. And wait. They try to pass...
the time. And wait. For something that doesn't come.
Yukon, the next-generation SQL Server, is perennially promised, but never delivered. Like Beckett's "Waiting for Godot," waiting for Yukon is becoming more and more like the Theater of the Absurd.
After being hyped and re-hyped for years now, after thousands of pages of ink (and electrons) have been spilled about its pros and cons, Microsoft announced last week another lengthy delay until Yukon's official release. Read our coverage of the story here:
- Security issues delay Yukon
- Microsoft: Yukon needs more testing
- Opinion: Microsoft delays are predictable
The new SQL Server won't be released now until 2005. Five years between upgrades is an eternity in this business. One would think that DBAs and developers would be upset at this news. Are they?
At last autumn's PASS conference in Seattle, I observed that a certain percentage of SQL Server geeks were indeed very excited about the new release, especially after test-driving the beta and hearing speaker after speaker extol its virtues. However, an even larger percentage was indifferent at best. Many already had their hands full with SQL Server 2000 and were dreading the upgrade process.
What's your opinion about Yukon's delay? Don't care? Crushed? Ecstatic? Do you think it's more important Microsoft release a robust and secure product than a glorfied beta? Email me at email@example.com with your thoughts.
Perhaps the best approach is a combination of hope and indifference, just like Beckett's Vladimir and Estragon.
Ed V. writes: "The issues of a stable or unstable product are important, but this is becoming a sideshow. The real issue is Microsoft's inability to deliver a stable product in a timely manner. The fact that Yukon looks so good compared to SQL Server 2000 is a major part of the problem. SQL2K is not the competition, and although Yukon looks good against 2002 versions of Oracle and DB2, by the time it is launched in 2005 Yukon will look outmoded. When looking at the upgrade path from SQL2K, the issues for many sites will be: 1) A lower cost move to outmoded technology that is unlikely to be refreshed for 3 - 5 years, and 2) An initially higher cost move to current technology that is likely to stay leading edge. Most businesses that expect to grow need technology that can help beat their competitors, and this could drive a significant number of customers away from SQL Server.
David R. writes: Most organisation run database systems from multiple vendors. Management perceived earlier versions of SQL Server as unsuitable for large critical systems. Largely due to the proliferation of smaller unmanaged applications that used SQL server. With SQL 2000 perceptions have changed. Applications have grown up and larger applications have also moved to SQL Server. Finally, organisations have started rewriting their legacy critical applications using SQL server. They have been told constantly about all these new features and are expecting to use them. The current Yukon delays will be noticed by management when the applications go live or are delayed. I am not looking forward to explaining why I wasted all their money on upgrade protection. I am sure the other database vendors will capitalise quite well on these issues. Microsoft has always been known for announcing and releasing products too early. With SQL 7 and 2000 they gained a lot of credibility with getting it right. Now it seems that they have gone too far and are waiting too long. My opinion is that the delays will close the door on a lot of redevelopment / migration to SQL opportunities. I also think that management will decide to stay with 2000 longer rather than upgrade. I am a strong SQL advocate, but I think my future will be brighter if I brush up my Oracle and DB2 skills now."
Gregory L. writes: Sure, I'm sorry to hear about the Yukon delay, but then again does this upset me? No, not really; let me tell you why.
I'm not sure if most shops are like ours, but we rarely jump out there and install the latest release of software. We do this for a number of reasons. The first and fore most reason is we want someone else to find the bugs in the software instead of us. When you don't have a lot of staff to manage SQL Server changes, we can't afford to spend extra effort chasing down vendor software problems. We would rather wait for the vendor to resolve some of these bugs, so basically we like to wait until the first SP is released if possible.
The phrase "If it ain't broken don't fix it" come mind regarding when I think of the second reason. Why would anyone need to install Yukon or any other version of SQL Server if you don't need any of the new features, just yet. Implementing new releases always presents challenges and sometimes causes existing applications to break. Normally these failures are due to some change in the software. So why have customers frustrated over a migration to a new version of SQL Server if it is really not necessary.
Lastly we are always concerned about money. If a new version causes you to buy new licenses and hardware then possibly the cost benefit is not there.
I'm willing to wait for Yukon. Hopefully this additional time delay will allow Microsoft to develop a more stable and secure product."
Doug S. writes: "Mostly, we use a rule that says "it's not ready for production until the first service pack is released". I feel that its far better to "get it right" instead of shipping something that's got some known problems."
Bharti M. writes: "Moving from one version to the next is a painful adjustment... The best analogy is to compare how a man starts a new relationship after a recent divorce... He is aware of the point of pitfalls however, there is little that he can do to prevent them, he simply takes them as they come. Similarly, Microsoft products are surely better than the previous but with each version the learning curve is incredible... The beta version is even worse mainly because you simply don't know what the breaking points are, and neither does the developer. Consequently, the dilemma with being in this industry is to keep on your toes with the latest and the greatest or else be out of a job."