You shouldn't necessarily need a function to convert a VARCHAR date to a datetime to save in a table... it's one of those "instrinsic conversions" that happen auto-magically if the...

KH did it right... if you expand upon the example KH provided, it works great, is nasty fast, and does not require a trip to the app.

Zia,

Instead of using COMPUTE, consider using WITH ROLLUP or WITH CUBE.  There's some great documentation on how to exploit the related GROUPING sub-totals for ROLLUP and CUBE in Books OnLine under...

Dave,

Very volitile subject... I got involved with a pretty big discussion on this subject before and I'll just say that, in all the "million-row" testing I did, I satisfied myself that properly...

They say that one test is worth a thousand calculations and a million guesses... Noeld is correct.  But don't take my word for it... I expanded my normally well behaved...

Very nice... very nice indeed...

These might execute a bit quicker... in fact, considering that they're single line formulas, you may not want to use them as functions, at all.. just...

For the data presented, this'll do it... and it doesn't assume that there will only be different 3 rows in the return...

SELECT Col1,Col2,Col3,MAX(Col4) AS Col4
FROM yourtable
GROUP BY Col1, Col2, Col3
ORDER...

For the data presented, this'll do it... and it doesn't assume that there will only be different 3 rows in the return...

SELECT Col1,Col2,Col3,MAX(Col4) AS Col4
FROM yourtable
GROUP BY Col1, Col2, Col3
ORDER...

Whether the column allows NULLs or not, the following will produce returns only if the column is NOT NULL and the column has something other than all blanks regardless of...

It seems like a pretty straight forward error message... your table won't allow nulls in the Status column but you're trying to add rows from data that contains no status. ...

I understand the question but why must it be without a sub-query?  Seems like a very unnecessary limitation unless you're doing this in an early version of MySQL or something...

Uhhg... I know you're right and it works great, but I just can't bring myself to do it  because way deep down inside, I know...

WOW!  Another myth destroyed.  I was always led to believe that BCP was always faster than BULK INSERT.  Phil's post

prompted me to do some long overdue testing on that fact. ...

Try this, Donald...

select * from xyz where datefield >= DATEADD(dd,DATEDIFF(dd,0,GETDATE()),0)

                          AND  datefield <  DATEADD(dd,DATEDIFF(dd,0,GETDATE()),0)+1

The DateAdd/DateDiff thing strips the time from the date without the expense of a character conversion and without...

Finally... the voice of reason...