Finally your turn to agree with me .

Bingo :

SELECT dbo.sysobjects.name AS TableName, dbo.syscolumns.name AS ColName, dbo.systypes.name AS DateType, dbo.systypes.xtype

FROM dbo.systypes INNER JOIN

...

Can't really help you without the structure and sample data. Have a look at DateDiff, between, , DateAdd in the books online for a start.

NP... don't waste another day next time .

k.. I just hope I understood what you needed .

Declare @A table (id int not null identity(1, 1), name varchar(50) not null primary key clustered)

Insert into @A (name) values ('atr')

Insert into @A (name) values ('d')

Insert into @A (name) values...

To see the rest of this thread :

http://www.sqlservercentral.com/forums/shwmessage.aspx?forumid=9&messageid=201273

That's an easy fix .

Declare @id

set @id = 0

Update dbo.table set @id = FakeIdent = @id + 1

Done .

Got it... change the in to inner join >>

SET SHOWPLAN_TEXT ON

GO

Select PkADP, ADPname from dbo.ADPS A INNER JOIN (Select EachId from dbo.fnSplit_Set...

Well this I'll have to find out another time.. thanx for the info.

They are talking about the function ISNULL not the operand "is null".

Here's another plan : Add the identity field directly to the main table to avoid generating it on the fly... You just saved yourself from a full table insert. ...

I would have to see the execution plan.. if it's not a cast then you might have to force the optimizer to choose another route. This is a case...

No offense, but this seems like a lot of overhead for such a simple task... Maybe it's just me but xml reminds me of 1950s flat file dbs... don't see...