What troubles me about this sort of discussion is the more general question; do we need to keep "improving" databases? Or maybe said another way, are these things really "improvements"? ...or are we really just adding more stuff to SQL Server so Microsoft can keep the revenue stream flowing for years to come.

If you work in an industry where you use SQL Server to do truly cool, modern, very important stuff - I say, great for you - but most people I know using SQL Server are doing the same mundane inventory tracking, record keeping, book balancing, web supporting stuff that I think 95% (maybe 99%?) of the world does with it.

It seems this concept of "improving" is great for a very small percentage of the SQL Server user family and this begs the question, if a company is "improving" something for 1%, 5% 10% of the user base - is that really any "improvement"?

I am also a bit peeved when talk of these "improvements" comes out of Microsoft itself, or those beholden to the company. Thats just a bass-ackwards way of defining improvements, and worse, judging them as good or bad.

Having survived Vista, Office 2007, and .NET I have seen that when Microsoft "improves" something they usually botch the job, over-complicating what was once easy, and then forcing the marketplace to accept what amounts to a step backward.

As a manager and a guy who has to account for the payroll budget, I dont see these things as improvements - they make the tasks twice as hard, twice as confusing and twice as long to complete. If that is what now passes for "improvements", well, it sure was nice of the Japanese to "improve" Pearl Harbor.