The Microsoft article uses a strangely anthropomorphic explanation, "Emptyfile assures you that no new data will be added to the file".  I'm happy to take DBCC's assurances any day but...

Hello Bevan,

In my haste I didn't try a null date, but you're right of course. Converting to varchar first means that no date or other type is presented to...

Thanks Bevan and John for the explanation. 

I had a chance to use it in a real scenario today but I hit a snag with Null values.     ...

That's a brilliant idea.  I often want to paste a small set of data into an email and it takes effort to make it readable even with SSMS's text output. ...

Excellent article - this technique is invaluable, especially for scheduling reports with varying periods.

Uncannily, after a long time thinking "there must be a better way to choose dates", I sat...

I have tried it now and you're right, Count(*) and Count(1) behave the same on version 2008 R2 (RTM) - 10.50.1600.1 (X64). I misread the QotD text.

..

Could this be a reason for preferring Count(1) over Count(*) - which is otherwise not an easy decision.

Edit: No indeed, I misread the question.

Does the Question of the Day of 2015-02-18 have a bearing on this discussion?

That is, COUNT(1) is the more reliable option because COUNT(*) will fail if the user is denied...

Did anyone mention that variables are also accepted by COUNT()?

DECLARE @i int

SELECT COUNT(@i) FROM [CountTest]

-- returns 0

DECLARE @i int = 1

SELECT...

Thanks Sean, the penny has dropped now.

Todd Reddinger (9/30/2014)

---

the ORDER BY in the OVER clause orders the records returned by the select. By adding an order by on the select, a differect set of records is...

The mistake in the question was overlooking the fact that sys.syscolumns will have different contents in different databases.

But more than that, you couldn't guarantee the statement would return the same...

At least following the recent spate of RAND() questions I found it quicker to spot that the question had gone wrong. I'm beginning to feel I finally know...

TomThomson (9/29/2014)

---

SQLRNNR (9/29/2014)

---

andrew.ing (9/29/2014)

---

Something I've never quite understood is why RAND() is like it is - that is, contrary to what perhaps most people would want and expect: creating a...