Hehe.

It's gonna be my turn to be shot down... gonna post my question in a new thread.

Come to think of it... that's gonna be only the 3rd thread I start...

Do you have something like this??

proc 1

insert into... exec proc

proc 2

do some stuff

call proc 3

proc 3

Select that inserts into proc 1

I've never heard of that problem... but that could be...

Did you add the missing rollback too?

If yes, then try removing it too see if it fails again.

HTH... this might sound strange... but I hope it fails again .

Maybe that could do it. Try making a new proc for the insert statement only. If that doesn't solve your problem, I'm officially out of ideas... or...

HTH.

I just wish that someone else with a more exact answer could post under this thread... Maybe someone from Teched?

Maybe I will extend the challenge... but Joe Celko's gonna kill me if I post code like that, or have me banished from sql...

Anyways, the problem with sysdepends is that...

Sorry for the confusion, I meant cariage return as in
in html. As I was trying to explain, it's just a matter of a characther being lost between...

The only "easy" way I see of doing this is at the client side. just takes 3 list and an IsInList function... not that hard to code and would...

Have you been able to solve your problem?

One more thing to add.. Exists stops executing its select as soon as a match is found, which is obviously not the case with a select: making it even faster...

I would expect that they had trouble with the @TableName.ColName syntaxe. The compiler probabely needs to assume that anything that begins with a @ is a parameter/variable unless it's...

Can you post the query, the execution plan and the table definition with its indexes... maybe we can speed it up still (don't know any time out settings in reporting...

Select will be slower than exists.

Exists returns true or false. A select will return a recordset, which is longer to create (especially if there are many rows in the...

Ya just relooked at that part, they are using the sp_depends stored proc to get that info... and we all know how reliable that info is.

Still seems like a great...