declare @s varchar(500), @x xml;

select@s='1,2,3,4,5',

@x = '' + REPLACE(@s, ',' , '') + '';

selecta.value('.', 'int') as s

from@x.nodes('/a') t(a)

edit: I see 'Pandian S' beat me to it with the XML hack/solution...

Cool. I want the time to see if I can work out what one of your 'improved' compression methods might be...

No more than running it so far - Real Life got partly in the way...

HA! So have you all seen the featured article[/url] today?

I think the stand out phrase for me is "I will be using the power of XML" :w00t:

Reminds me of:

I've never seen anything to suggest that is necessary, no.

My favourite article for this stuff on here is http://www.sqlservercentral.com/articles/SQL+Server+2005+-+Security/sql2005symmetricencryption/2291/

There are plenty though - search the tags if you are interested.

Paul

Just in case there is any doubt at all about this: If you specify a file group for a clustered index (primary key or unique constraint) in a CREATE TABLE...

Gift Peddie (6/24/2009)

---

Actually in ANSI SQL 1999 Subquery is equal to an INNER JOIN so it is techincally faster than OUTER JOIN because of the OUTER JOIN default NULL. ...

That would work, but I have a feeling Flo is about to delight and amaze everyone with a rather nifty running-totals implementation.

Just a guess :w00t:

Generally speaking, the query optimizer doesn't much care how you write your queries.

Ok, so that is a bold statement - but it is at least partly true, at least to...

DCPeterson (6/24/2009)

---

Keep in mind that you don't need to include the clustered index key column(s) in the included columns list because the clustered index key is automatically part of any...

Carl Federl (6/24/2009)

---

I think if you look further, you will find Cores 3 and 7 are handling all of the network IO , hence the higher CPU usage. There...

Cool.

Though the two queries are not exactly the same - check the query plans for both 😉

Also, I assumed you wanted the student ids concatenated in ascending order - you...

Could I ask also that you confirm that we should compute the running totals in descending percentage order?

For example, if the available percentages were 35%, 20%, 15%, 10% and 5%...

A nonclustered index on ACCT_DEBTOR with OCCURRENCE and LANDLINE_CONTACT_NO as included columns, I think.

(I would also be naturally inclined to create a clustered...

CREATE TABLE #Class (row_id INT IDENTITY PRIMARY KEY, class_id INT NOT NULL, student_id INT NOT NULL);

CREATE TABLE #Classes (class_id INT PRIMARY KEY, student_list VARCHAR(8000) NOT NULL);

-- Test Data

INSERT#Class (class_id, student_id)...