How did you get your result?
I used the following query and get other numbers:
WITH myBeers (<see CTE in the question> )
SELECT ISNULL(bt.OwedTo, bb.OwedBy) who, SUM(bt.Beer)...

To be honest, I did not even understand the question and felt like a newbie...

I'd prefer to use
FIRST_VALUE(pts) OVER (Partition by Team ORDER BY Year ROWS BETWEEN UNBOUNDED PRECEDING AND CURRENT ROW)
since this performs better (does not spill to...

I guess they needed the 29 days to remove those silly clause, that everybody needs a vacuum cleaner in the basement, which was added by me in my youth while...

I doubt, that this will be fast. It does not use an explicit cursor, but practically does the same (or something worser) with its while loop.

If someone really...

Wouldn't the minimal privileges be the NTFS permissions to delete the ERRORLOG.? files? Sure - you can't replace / modify the current ERRORLOG file (SQL Server holds its tumbs on...

I wonder, that the only connect item I found for a column based MIN/MAX (similar to Oracle's LEAST()/GREATEST()) got only 4 quotes since last year (https://connect.microsoft.com/SQLServer/Feedback/Details/3085573).

Maybe someone...

Why do you not use the build in ROUND function?
It takes three parameters:
- the value
- the rounding length (use negative numbers to round to a power...

my apology for the confusion - I tested it only on SQL 2014 (since I was to lazy to set up a new 2016 server just for testing and did...

On which version did you run your test? On SQL 2014 SP2 Dev. the query will only return one row (and I stumbled over a few scripts which failed, since...

Another question.
If I understood it correct, it has to read all segments / pages for the city column, when I execute a SELECT COUNT(*) from customers...

Thank you very much for the quick answer. I always wondered, how exactly SQL Server does this stuff and your article (plus the answer) makes it a lot more clear...

any idea, if (and if yes then how) small datatypes as TINYINT (= 1 byte) could be compressed in columnstored indexes?

Its clear, that I will save space...

You should use an INLINE TABLE VALUE fuction instead of a scalar function - this would save a lot of time when you import many datasets

CREATE...

Short remark: a FILLFACTOR = 90 on a Identity Column (= auto increment) in a read-only-table (at least no space-growing varchars) makes no sense and wastes 10 % space. Even...