Thanks for the recommendation Cold Coffee 😎

Azeem, if you have any questions after reading both parts of the article and examining the code samples provided there, please post them here...

skcadavre (6/8/2010)

---

Please don't message me about specific problems, I get a notification for threads I've posted in so if I decided not to reply there was a reason.

I answer questions...

Hey,

Thanks for that. Rather than debate how the statistics are being used, let's take a look at the expression you are asking the optimiser to plan for. Both...

And the variance would be 10? :unsure:

What is the earliest month for person 3 in your original sample data?

Is it month 1 or month 11?

Month 12 is not in the list.

*** never mind ***

That's...

If you find yourself needing the date-only portion frequently, it might make sense to pre-compute it.

You can minimise...

Assuming that the current date acts as the cut-off for deciding whether a month number belongs to last year or this year...

DECLARE @Sample

TABLE (

...

Excellent question, Hugo.

The simple answer is that the query is a poor one, and SQL Server 2005 happens to come up with a less disastrous execution plan than 2008. Try writing...

CirquedeSQLeil (6/7/2010)

---

...Procs sometimes go through me - and that is getting better. It is a slow process of education. Get the dev team up to speed enough that...

Steve Jones - Editor (6/5/2010)

---

In a source control session at SQL Saturday #22. Surprising that quite a few people don't use an VSC for their SQL Code. What do...

sgalbra (6/4/2010)

---

It is the exact same database, and the exact same compatibility level (90). After the upgrade, I rebuilt indexes, updated statistics (even for the underlying tables), and updated...

Why do you say that? Another mistake?