It's...

John,

That link seems to be broken, which is a shame since I would like to be one of the cool kids that's in on the joke. :unsure:

Paul

Either method will work fine, though the Tally solution is more flexible and generally performs a little faster. Some DBAs would prefer the UNPIVOT solution for its economical syntax....

Just to add an alternative, if the number of columns is fixed:

DECLARE @TABLE_FROM

TABLE (

A1 VARCHAR(100),

...

DELETE

FROM @t1

WHERE NOT EXISTS

(

SELECT *

...

Without duplicating the expression:

DECLARE @temp

TABLE (

id INTEGER NOT NULL,

phn INTEGER...

Yes, but your solution is documented and supported 😎

If you're not authorized to change anything, it can't be changed :laugh:

Do it application-side, or explain yourself clearly, and provide a valid reason for doing this.

Paul

It is possible to write any JOIN as the equivalent sub-query. The reverse is not true, however.

SQL Server doesn't care too much...

If this is new development, you should avoid NTEXT it is deprecated, and will be removed from SQL Server. If you need to store large Unicode strings (> 4,000...

One other small thing: PRIMARY KEY is CLUSTERED by default, UNIQUE constraints are NONCLUSTERED by default.

Paul

Seems an odd requirement, and the wrong place to do it as Lutz mentions.

Nevertheless...

DECLARE @demo

TABLE (

data ...

Yes...

David,

If I ever end up having to maintain that code, you may expect a cold stare or two in the hallway 😀

Paul

Hey Steve,

Good code - but if you wrap it in [ code="sql" ] [ /code ] tags (remove those extra spaces), it will look much nicer on the page 😉

The...