Exactly: the orphaned ones will not show up in the GUI, so you can't delete them there, but you can DROP them using the T-SQL commands.

Please include the exact sp_dropuser command you used and the results of SELECT * FROM sys.database_principals as well.

The user account can't be used to access data.

But you should be able to drop the user account from the db.

What error message do you get when you try to...

Hmm, that's a difficult issue. With what you're doing that won't be easy.

I thought earlier about creating a full-text index across the "notes" column. In theory that could...

You can't drop the user account(s)? [not the actual login]

You can refer to previously defined CTEs in a later CTE, but you must reference it like a table, i.e. JOIN to it, since that is effectively what it is.

Did the TRUNCATE TABLE actually work? There are, of course, certain restrictions to using TRUNCATE.

Did something else in the batch fail and cause a rollback? Unlike Oracle, a...

Yep, you need to verify that the log will actually shrink as much as want before issuing the SHRINK command. That means putting the db in simple mode first.

You...

Get the free SQLIO tool and gather stats from that too. That will give you a best-case baseline for disk io (and, if the numbers aren't good/right, allow you...

You should be able to use DROP USER or sp_dropuser to get rid of users in the db, whether they are linked to a login or not.

INSERT INTO foundTable ( note_#, product_code )

SELECT n.note_#, p.product_code

FROM notesTable n

INNER JOIN productsTable p ON

n.notes LIKE '%' + p.product_code + '%'

...

Any constraints will be verified before your trigger ever gets control. Thus, you cannot handle those types of errors in a trigger.

In general, you use BEGIN TRY ... END...

Please post the valid formatting chars the existing function accepts and the expected output from sample formats.

What specifically does that mean? I'm guessing the # is zero-suppressed...

You don't actually need PATINDEX for that, LIKE will work just as well:

WHERE string LIKE ' [^ ]%' OR string LIKE '%[^ ] [^ ]%'

To paraphrase Clinton, it depends on the definition of "work". If two people have the same total,...