Viewing 15 posts - 6,211 through 6,225 (of 8,416 total)
Quassnoi,
At the risk of interrupting your flow here, are you aware that you are missing something very important from your 'group-wise min/max (with ties) discussion?
There is a QO transformation, available...
February 6, 2010 at 6:54 am
Adam Machanic (2/5/2010)
February 6, 2010 at 3:32 am
RBarryYoung (2/6/2010)
February 6, 2010 at 2:48 am
TheSQLGuru (2/5/2010)
February 5, 2010 at 9:20 am
Clearly a bug in the optimizer - take it up with the SQL Team :laugh: :laugh: :laugh: :laugh: :laugh:
No, seriously, I got my caveat post in first, so AHA! back...
February 5, 2010 at 9:19 am
I opened a Connect suggestion of my own, for CLR user-defined analytic functions (streaming multiple rows input and output thingies :-D).
I would hope to encourage some casual readers of this...
February 5, 2010 at 9:16 am
TheSQLGuru (2/5/2010)
Quite interesting Paul! I have seen numerous real-world systems where that was a killer. Things that make you go "HMMMMM"!!
Well yes, absolutely. I'm not arguing the...
February 5, 2010 at 9:14 am
Adam Machanic (2/5/2010)
February 5, 2010 at 9:06 am
Adam Machanic (2/5/2010)
February 5, 2010 at 9:03 am
Adam Machanic (2/5/2010)
February 5, 2010 at 8:38 am
Adam Machanic (2/5/2010)
Might want to re-read what Bob Dorr has to say about this technique:
😀
In Bob's case he was only working with...
February 5, 2010 at 8:37 am
TheSQLGuru (2/5/2010)
February 5, 2010 at 5:38 am
Pedro DeRose [MSFT] (1/22/2010)
Which, I think, may be a good slogan for SQLCLR: better than cutting boards with a hammer. 🙂
Do you mind if I adopt that gem into...
February 5, 2010 at 4:18 am
Adam Machanic (2/3/2010)
February 5, 2010 at 4:16 am
d.majoor (2/5/2010)
Large tables would require a lot of column typing though.
You could use SELECT DISTINCT * to overcome that (instead of using the group by).
February 5, 2010 at 1:51 am
Viewing 15 posts - 6,211 through 6,225 (of 8,416 total)