Quite frankly: If you think you have to know this, you're very likely doing something wrong. Still, I've earned yet another very arcane bragging point.:hehe:

I'm currently hiring. I'd like to know which 9% selected the CHTM option:hehe:

I think the only thing I take away from this question is confirmation that dynamic SQL must be avoided, but if you need it you should at least name your...

Interesting question. Too bad none of the answers are correct. BI edition does not support buffer pool extension.

Oh, and your reference link doesn't work.

http://msdn.microsoft.com/en-us/library/cc645993(v=sql.120).aspx

If SOUNDEX() as implemented in SQL Server was remotely useful outside of English-speaking countries I might care one way or the other. As it stands it's one of the quaint...

Ford Fairlane (11/7/2013)

---

Very good question - got caught out with as well. Should have realized that with out the where clause that each max() is evaluated individually... nice to learn...

Well, good question, but the correct alternative was a bit misleading. You don't get a zero back, you get an empty result set. That's a different answer.

Nice job linking to the Russian language online docs...

However, there is nothing in that text which says anything about whether ALTER TRIGGER affects the enable/disable state of the trigger.

Hugo Kornelis (9/3/2013)

---

You must have overlooked the things already mentioned about ordering by Rune Irvin, Terreador,...

Interesting. Those are somewhat different from the SQL 2012 plans. And goes to show that when something isn't documented it may very well change between releases without any formal notice.

L' Eomot Inversé (9/3/2013)

---

Rune Bivrin (9/3/2013)

---

Just look at the execution plan and tell me if there is no difference.

Or even try this, where NEWID() is completely eliminated from the second...

John Mitchell-245523 (9/3/2013)

---

There's very little difference. On SQL Server 2008 R2 SP2, there was...

Just look at the execution plan and tell me if there is no difference.

Or even try this, where NEWID() is completely eliminated from the second query.

-- 1

SELECT ROW_NUMBER() OVER (ORDER...

Great question!

I've never used DQS, but the correct answer just seemed the most logical.

Got it. But to be precise there won't even be a result set. Not even one with zero rows.