Bob Hovious (4/13/2009)

---

Gus, I rewrote your function as an inline table valued function and it gets the job done in half the time of the ConcatMaxTest function, but still three...

Jeff Moden (4/13/2009)

---

I can almost guarantee that there will always be 50,000 distinct SomeInts on the million row test. And, although the SomeLetters2 column is pretty random, the average...

Jeff Moden (4/13/2009)

---

GSquared (4/13/2009)

---

CLR does a faster running total than the quirky update too.

As we've known for quite some time, SQL just plain isn't good at what it isn't designed...

You're on the right track, but what you need is "Case When Else". Look up Case in Books Online, it's got samples, syntax, etc.

CLR does a faster running total than the quirky update too.

As we've known for quite some time, SQL just plain isn't good at what it isn't designed for. It's...

You can't create a view on tables that don't exist. Won't work. You need to create the tables first, then create the view.

Just for one more test, I took the type and value functions out of the XML version of the UDF, and that took it down to 6.5 seconds CPU and...

Just to test it a bit further:

CREATE FUNCTION dbo.ConcatXMLTest (@SomeInt_in INT)

RETURNS VARCHAR(8000)

AS BEGIN

RETURN (SELECT

...

Jeff, in your posted test harness, you need to rename the functions in the final tests. It has "SELECT SomeInt, dbo.ConcatTest...", but it needs to be Concat8kTest and ConcatMaxTest,...

GilaMonster (4/11/2009)

---

I do, solely for error handling in SQL 2000. Since there isn't a TRY-CATCH construct...

Ruby:

What I'm going to suggest is use a simple split, then select all rows where you end up with a space in the first or last name, after it's been...

Todd:

I just looked at the name parser at SQLTeam that you linked to. It's a useless toy. You have to enter the format codes for each name. ...

Barry:

Where did you get using 0 in parsename? I'd never seen that before, and it's not in BOL for 2k5, so I tried this:

select parsename('gus.gwynne',0)

Result is null. Put...

Try using this to delete dupes. It's MUCH faster.

;with Dupes (Row) as

(select

row_number() over

(partition by Project,PLC,CP_Bill_Hours

order by Project,PLC,CP_Bill_Hours)

from #temp_Report2)

delete from Dupes

where Row > 1;

This assumes you're using...

What error message does the second problem give you?