One of the coolest thing about being an MVP is hanging out with people who are smarter than yourself.  As evidence of that reality, I have a little snippet of a discussion here between a couple of my MVP colleagues that I found both interesting, funny, and edifying.

Steve Kass: The year 2005 will be one second longer than a usual non-leap year.   For the first time since 1998, the IERS has announced that the earth's rotation is running slow relative to atomic time standards, and the last minute of December 31 will contain 61 seconds in order to get things back on track.  The announcement is posted at ftp://hpiers.obspm.fr/iers/bul/bulc/bulletinc.dat.

<?xml:namespace prefix = o ns = "urn:schemas-microsoft-com:office:office" />

 

Kevin (to myself): Hmmm, interesting.  But how does this impact SQL Server?

 

Steve:  All versions of SQL Server don’t support valid DATETIME values such as '2005-12-31T23:59:60.500' that occur during leap seconds.  In addition, SQL Server's DATEDIFF function does not take into account leap seconds.  Be prepared!

 

<?xml:namespace prefix = st1 ns = "urn:schemas-microsoft-com:office:smarttags" />Hilary Cotter: So when I'm in Times Square counting down the seconds to the New Year, how do I count  -  10,9, 8, 7, 6, 5, 4, 3, 2, 1, 1 Happy New Year?

 

Steve:  Good question.  Apparently the answer is no.  The leap second will be added to the last minute of the UTC year, so you will only need to count with a stutter if you are in

Trafalgar Square
, not Times Square.  When the countdown gets underway in Times Square, the leap second will already have come and gone.

I’m curious: Do any readers out there have such a time-sensitive application that a single second will make a difference in your app?

Steve keeps a blog at http://www.users.drew.edu/skass/sql/ that is well worth your time.

Hilary has written the best book around on SQL Server replication.  Get details at http://www.nwsu.com/0974973602.html.


Cheers,

-Kevin