There is a comment which deals with NULL and NOT EXISTS.

Read the comments in the link I posted earlier.

You can decide for yourself and make your mind up be reading about normalization rules

http://www.datamodel.org/NormalizationRules.html

You can use FOR XML EXPLICIT

His index problems are answered here

http://www.sqlteam.com/forums/topic.asp?TOPIC_ID=129182

See http://weblogs.sqlteam.com/peterl/archive/2009/03/19/NULL-NOT-IN-conundrum.aspx

The "trick" is the INTEGER division made up by "/ 7 * 7"

The first division divides by seven and discards the fractional part.

Then the integer part is multiplied by seven.

And if you also want to reuse the identity values, see

http://weblogs.sqlteam.com/peterl/archive/2009/03/10/How-to-efficiently-reuse-gaps-in-identity-column.aspx

Another reason to use COALESCE instead.

DECLARE@num NUMERIC(15, 0),

@dt DATETIME

SELECT@num = 20090616052600,

@dt = '20090316 22:27:20'

SELECTDATEDIFF(HOUR, '20090316 22:27:20', '20090616 05:26:00'),

@num,

@dt,

DATEDIFF(HOUR, @dt, STUFF(LEFT(@num, 10), 9, 0 ,' ') + ':0')

Is it possible for you to attach the execution plan in "graphical" or xml ?

The *.sqlplan file will do!

From what I can see the two executions plans are different...

I have seen various execution plans depending on the order of tables in the FROM part.

In most cases, there will be no difference because SQL Server rearranges them in an...

Also see

http://www.sqlteam.com/forums/topic.asp?TOPIC_ID=77311

Another 4 page rant over here

http://www.dbforums.com/microsoft-sql-server/1644506-rant-referential-integrity-not-possible-sql-server.html

If the CHECKSUM is deemed as acceptable function for hashing out values (1 chance in 16,777,215), using this code gives one chance in 3,047,466,240 to get a duplicate value!

DECLARE @Codes...

You're welcome.