I have just arrived on Eileen Brown's team as the Data IT Pro evangelist. During my extensive  interview process I was asked by Keith Burns about the common language runtime (CLR) in SQL server 2005 and I replied with a quotation from his blog "that if it can't be done in SQL it isn't worth doing".  That would seem to rule out blogging but with the arrival of SQL server 2008 and it's ability to handle the vast amounts of unstructured data we encounter both at home and work this could well turn out to be absolutely true!

The other point about this remark is our digital immortality.  At our internal technical conference last week, one of the speakers talked about the first computer he had to use, a secondhand missile guidance system which he had to hook up to a mass spectrometer. After spending many sleepless nights getting the system working he prepared a comprehensive demo for the chemistry lab staff.  They weren't impressed at all by his technical wizadry and just wanted to get on and use the mass spectrometer. So he slightly 'modified' the code to make the machine crash randomly and added a lot of extra unnecessary steps to the startup instruction manual than was necessary.  This all came back to haunt him when the machine kept hanging as his random number generator turned out not to be so random and he had to use his old time consuming instructions to continually restart the system in order to debug his code. 

So not only are your finest moments preserved forever but also our worst mistakes.  Which needs me neatly onto one of the darker corners of SQL Server and that is deprecation of features as occasionally we need to tidy up some of the the throwbacks to earlier versions of the product.

There are two parts to this - features that won't be supported in SQL server 2008 as shown below:

And features that will be deprecated in future releases SQL server after the 2008 version.

http://msdn2.microsoft.com/en-us/library/ms143729.aspx

So careful planning is needed if you plan to migrate to SQL server 2008.