I agree... although your intentions are good, I'm sure, this is nothing more than spamming a forum.

I have found that most transactions are simply not necessary... for example, if your transactions have no ROLLBACK code, there is no need for the transaction.  The WITH (NOLOCK) suggestion...

Agreed... a little more info would help... should it be a date, and sequential ID, what???

Rodger,

Did anyone change the default datatime format on the non-production box?

This is why you should not trust defaults... things change. 

Perhaps, if you posted the code, we may be able...

Keith,

Yes, there's an easy way and it does NOT require the overhead of a cursor...

Substitute the name of yourtable for "yourtable" everywhere in the code below...

First, make a function...

I', a bit confused... you know the number should always come in having 5 decimal places but you want to convert VARCHAR to INT and you ultimately want to "do the...

Try not using a cursor ...

Thanks, Bill,

I do have to emphasize that performance will usually be better if you do this simple type of thing in the code mainstream instead of a function.  Functions are...

Jules,

The two functions I wrote do get rid of the time portion and they do it with only trwo functions instead of the 5 you used.  AND, they did it without...

My recommendation would be not to make a trip to a character based date and back again...  I'd also not make a stored procedure for it.  I'd likely do it...

Dunno about VB datetimes but in SQL Server 2000...

I think you'll find that 23:59:59.999 will round up to precisely 00:00:00.000 and that 23:59:59.998 will round down to 23:59:59.997.

SELECT CAST('23:59:59.999' AS DATETIME)

SELECT CAST('23:59:59.998' AS...

I agree... a decent article... but it DOES sound like an advertisement and provides no real info as to how to do the sizing estimate if you don't have those...

Not a problem, Nick.  You had the right idea...

In SQL Server (very cool and very unlike Oracle), there are two tables that are temporarily formed (in TempDB) and available in...

Nicely done... a couple of minor things you may want to change...

1. If t2.ID>t1.ID then t2.ID cannot be equal to t1.ID.  Remove the t1.ID<>t2.ID criteria.
2. In case the amount of time is...

I haven't checked your logic but the problem with the triggers is that they are written in a RBAR (pronounced "ree-bar", my less than affectionate term for "Row By Agonizing...