These kind of queries have real usage, for example if you want to preserve "duplicate rows" of Col1, Col2, which cannot be preserved if you only Group By Col1

Following workaround might be of interest for others (this works in both 2008 & 2012):

CREATE TABLE #t_test (i INT DEFAULT 1)

SELECTc.COLUMN_NAME, c.COLUMN_DEFAULT

FROMtempdb..syscolumns sc (NOLOCK)

INNER JOIN tempdb..sysobjects so (NOLOCK)

ONso.id = sc.id

INNER...

Took a while to reply to this, but I tried your method and still wasn't any difference as far as I could see.

Gave up on this, one just have to...

sys.index_columns view is your friend, it list regular and included columns for an index:

SELECTOBJECT_NAME(sx.object_id) AS table_name

,si.name AS index_name

,sc.name AS column_name

,sx.is_included_column AS is_included

,sc.*

FROMsys.index_columns sx

INNER JOIN sysindexes si

ONsi.id = sx.object_id

ANDsi.indid = sx.index_id

INNER...

The real question is, do you really need to import 4k of files every day, because that's pretty impressive customer flow :hehe: