Nice question.

At first I thought it would have been better to make this number 1 in the sequence, and make number 1 number 2, but seeing the...

Andrew G (1/13/2015)

---

He has a point, the first link from the answer states:

http://msdn.microsoft.com/en-AU/library/ms189838.aspx

Valid data types for an identity column are any data types of the integer data type category,...

easy one.

But the explanation is wrong - NULL values aren't counted, only rows are counted and a row can't be NULL.

Match

Nice question. Pity about the typing error, though ("." for ",").

julienchappel 38298 (1/13/2015)

---

Hi Steve this question of IDENTITY column data type your answer is wrong. The correct answer is quoted from http://msdn.microsoft.com/en-AU/library/ms189838.aspx:

IDENTITY (Function) (Transact-SQL) data type

Valid data...

Hugo Kornelis (1/12/2015)

---

I never liked implicit transaction, and I never will.

I am also glad that Carlo used PRINT, not SELECT. I thought that using SELECT would have started the implicit...

Nice question.

I think that having BEGIN TRANSACTION increment trancount by 2 instead of 1 is utterly silly. It's a really nasty trap that can cause pointless and damaging rollbacks.

eric.notheisen (1/8/2015)

---

I have had concerns about storage as a service for some time. I remember doing technical sales of Microsoft CRM against SalesForce.com years ago. My primary selling...

legeboka (1/8/2015)

---

Hi,

I’m trying to write a query that will select closest value to Amount field in #TestValue without go over #SM.GMAdj value

Outcome of the query should be something like...

The change to look a prices requires only very small change to Luis' solution for the version involving validity, like this:-

WITH cteMP AS(

SELECT MachineFk,

...

Good question.

But the first option wasn't possible even if ERROR_MESSAGE() could function outside a catch block, because if it did function it would produce the error message ("Cannot insert the...

dark

Phil Parkin (1/7/2015)

---

If something is non-deterministic, then it's very possible for the re-evaluations of the object test to miss all of the WHEN's and produce a NULL if there...

COUNT isn't the only one, because COUNT_BIG behaves the same way as COUNT. Also, like others who have already commented, I find it hard to see how it is...