Viewing 15 posts - 571 through 585 (of 708 total)
Yelena: That's a typo from early documentation that hasn't been corrected everywhere.
Indexed views may be used in SQL 2000 Standard Edition by specifying the NOEXPAND index hint. All editions allow...
July 5, 2006 at 3:08 pm
Looks like you can dodge all the CASE statements With a NullIf(value, 2):
NullIf(Min(IsNull(CONVERT(smallint, PermissionField1), 2)), 2) AS NewPermissionField1,
Testing (in results,PermissionField1 should...
July 3, 2006 at 1:11 pm
PIVOT works best when you are using aggregations (count(), sum() avg(), etc.). In this case, because you want to return every value instead of the results of an aggregation over...
July 2, 2006 at 8:54 pm
A really solid book is Ken England's Microsoft SQL Server 2000 Performance Optimization and Tuning Handbook. It's less than 400 pages, and packed with info. It's one of those rare...
June 27, 2006 at 11:14 am
This sounds like a homework question...
Have you tried it? What was the result and what would you like explained?
-Eddie
June 27, 2006 at 8:29 am
56,000 bytes. (call it ~100kB, depending on your fill factor)
If this hypothetical table were real, then you've already spent more of your company's/customer's money thinking about this than the cost of...
June 27, 2006 at 8:18 am
The trick that's helped me the most: just book the test.
Once I start getting comfortable with what will be on the test - but long before I'm ready - I'll register for...
June 27, 2006 at 12:06 am
Jeff: CREATE and DROP prohibit use of the database name. They must be executed in the context of the target database.
-Eddie
June 26, 2006 at 11:23 pm
If you wish to cascade some of the time, based on the situation, then it sounds more like a business rule than a data integrity rule. Business logic should be...
June 26, 2006 at 3:04 pm
Be sure to fire a USE statement first, then the database context is set:
USE
MyDB
GO
SET
ANSI_NULLS,
June 26, 2006 at 9:22 am
Every object has a four-part name: Server.Database.Owner.ObjectName
SQL Server will fill in the blanks with the current context if you leave any of them off. If I'm logged in to
June 26, 2006 at 7:20 am
It's hard to fight raw data, your plan of comparing reads and usage should be enough. If that doesn't make sense to the developers, then they are clearly stating that...
June 23, 2006 at 11:49 am
Just riding Ryan's solution, here.
To get yymmdd use 12 in the convert function, and 112 to get yyyymmdd:
SELECT convert(varchar, DATEADD(ss, @i, '01-01-1970'), 112) as 'My date yyyy'
June 23, 2006 at 10:48 am
What you describe can be done, but (and it's a little late I guess) structured naming solves a lot of the problem up front, grouping procedures by logical area, using...
June 23, 2006 at 10:41 am
>> 1. Is there a pattern to the ID list, or is it just a random blast of IDs?
> Random.
Shoot. Well, there goes that idea...
> Are there any obvious...
June 22, 2006 at 1:54 pm
Viewing 15 posts - 571 through 585 (of 708 total)