Also it could be the statistics are changing in your data. Make sure that they're up to date. It's really important, as noted above, to look at the execution plan,...

Database design theory, I'll leave to someone else. I'm using an ancient text called "Handbook of Relational Database Design" by Fleming & von Halle. It's too basic for most of...

Oops. Missed that one. I hate when I get hung on the first error I see.

How about

OUTPUT INSERTED.CategoryId

because there wasn't a column in the list called ID

I don't think that's possible. Unless there's a way to make the Compact database a linked server, there's no way to open a connection within TSQL. You could write CLR...

I agree that NOT using a cursor is an even better approach.

However, to answer your question, I wouldn't use the WHILE loop with a BREAK statement because it's really a...

It sure sounds like you need to set up Profiler. That will capture the stored procedure calls along with the parameters passed, no extra code on your part required. You...

A teenager down at the dojo says I'm losing my touch... He's testing in August. I'll "touch" him then.

Anyway, glad to be of use. One other point worth mentioning is...

Not even, don't worry about it.

You've got a lot of options. You mentioned triggers. That is one possibility. Another would be to use the OUTPUT clause from your data modification...

See the post above by Julius. That shows how to unpack the XML files.

Can you edit the sample XML to remove the angled brackets? I usually use square brackets and comment the code. The angled brackets results in blank data as you see...

And the place you want to go to look at these locks as they occur is sys.dm_tran_locks (p. 346 😛 ).

The reason you're getting an error saying those objects don't exist is because those objects don't exist in 7.0.

I actually don't have any more copies of 7.0 installed around here...

Also, from the sounds of things, you're doing RBAR (row-by-agonizing-row) processing. Can't you pass the ID's to your query as a set and then do a join against your data...

The order of locking depends on the query being run. That's why two different queries can lock in different order leading to deadlocks. You can usually spec it out in...