Jeff Moden (11/27/2012)

---

I've seen all of his arguments and appreciate them but they still boil...

Try this

Select event_id, event_ts, DATEADD(minute,(1+DATEDIFF(minute, 0, event_ts)/5)*5, 0) AS time_stamp_group

From cte_time

Looks like a typo

CROSS APPLY T.CommentsData.nodes('/comments_data/Row/id/from') AS S(N)

should be

CROSS APPLY T.CommentsData.nodes('/comments_data/Row/from') AS S(N)

You can do this in a single statement using MERGE

MERGE fk_test AS target

USING (SELECT 5) AS source(eno)

ON (source.eno IN (target.eno,target.fno))

WHEN MATCHED AND source.eno = target.eno THEN

DELETE

WHEN MATCHED AND source.eno =...

change

select @VXML.modify

to

set @VXML.modify

Assuming you have a column SeqNum to order the rows by, you can do this

UPDATE t1

SET Admit_Date = ca.Admit_Date,

Disch_Date = ca.Disch_Date,

...

Try this

SELECT x.r.value('(Contact/ContactID/text())[1]','VARCHAR(40)') AS ContactID,

x.r.value('(Contact/Name/text())[1]','VARCHAR(200)') AS Name

FROM @Xml.nodes('/Response/Invoices/Invoice') AS x(r)

You're welcome. Thanks for the feedback.

WITH CTE1 AS (

SELECT [ProfileId],

x.r.value('.','INT') AS id

FROM [Profile]

CROSS APPLY RelationshipTypeIDs.nodes('RelationshipTypeIDs/id') AS x(r)),

CTE2 AS (

SELECT x.r.value('.','INT') AS id

FROM @RelationshipTypeIDs.nodes('RelationshipTypeIDs/id') AS x(r)

)

SELECT DISTINCT c1.[ProfileId]

FROM CTE1...

About to be deprecated? Have you got a link for that?

If your talking about this connect issue, it's already...

dwain.c (11/21/2012)

---

Prettiness and speed are apparently not conflicting requirements:

PRINT 'Dwain.C COLLATE + CROSS APPLY'

SET STATISTICS TIME ON

SELECT @Email=RIGHT(

STUFF(a.email

...

Modded version of Dwains, much quicker

PRINT 'Dwain.C COLLATE'

SET STATISTICS TIME ON

SELECT @Email=RIGHT(

STUFF(email

,CHARINDEX('@', email COLLATE Latin1_General_BIN2)

...

Jeff Moden (11/19/2012)

---

@mark-3,

That was also pretty quick. Have you been using such a thing? If so, what's your take on performance here (And, yes, I still do the...

Jeff Moden (11/19/2012)

---

Eugene Elutin (11/19/2012)

---

Jeff Moden (11/17/2012)

---

Raymond van Laake (11/16/2012)

---

...

Regular expressions may be the fastest in another language but they're not necessarily the fastest in T-SQL because of the bit...

Something like this?

SELECT DB_NAME(database_id) AS DatabaseName,

SUM(CASE WHEN type=0 THEN size ELSE 0 END) AS MdfSize,

SUM(CASE WHEN...