Viewing 15 posts - 256 through 270 (of 582 total)
Google: TSQL split string
There are loads of threads about stuff like this.
Have you considered using XML, or not doing this at all(!)?
I hope you aren't using a production (rather...
Tim Wilkinson
"If it doesn't work in practice, you're using the wrong theory"
- Immanuel Kant
June 20, 2006 at 8:56 am
Tim Wilkinson
"If it doesn't work in practice, you're using the wrong theory"
- Immanuel Kant
June 20, 2006 at 8:48 am
Tim Wilkinson
"If it doesn't work in practice, you're using the wrong theory"
- Immanuel Kant
June 20, 2006 at 8:37 am
>Bigint is 8 byte, not 16.
ah yes, good point. I thought those mushrooms were a funny shape.
>Varchar can NEVER be more effective than numeric. It uses only 26...
Tim Wilkinson
"If it doesn't work in practice, you're using the wrong theory"
- Immanuel Kant
June 20, 2006 at 7:30 am
I remember in v6.5 in which quoted_identifier was off by default, I used to use single quotes for the outermost layer, then doubles for embedded quotes. So...
Tim Wilkinson
"If it doesn't work in practice, you're using the wrong theory"
- Immanuel Kant
June 20, 2006 at 6:32 am
And you have to divide the datalength() output by 2 for unicode (nchar etc.) strings.
Tim Wilkinson
"If it doesn't work in practice, you're using the wrong theory"
- Immanuel Kant
June 20, 2006 at 6:13 am
That won't help you.
BTW, are you still there?
Tim Wilkinson
"If it doesn't work in practice, you're using the wrong theory"
- Immanuel Kant
June 20, 2006 at 6:11 am
Can you post the individual statements, with their WHERE clauses, and the select from the view, with its WHERE clause? There shouldn't be any significant difference in execution times.
Tim Wilkinson
"If it doesn't work in practice, you're using the wrong theory"
- Immanuel Kant
June 19, 2006 at 11:35 am
This is unequivocally better than the solution those MS types came up with:
Tim Wilkinson
"If it doesn't work in practice, you're using the wrong theory"
- Immanuel Kant
June 19, 2006 at 11:15 am
Tim Wilkinson
"If it doesn't work in practice, you're using the wrong theory"
- Immanuel Kant
June 19, 2006 at 10:59 am
Yes, I believe CROSS APPLY does it but you force row-by-row processing of the outer recordset, so you might want to consider using joins instead if possible. If your UDF is an...
Tim Wilkinson
"If it doesn't work in practice, you're using the wrong theory"
- Immanuel Kant
June 19, 2006 at 8:36 am
>why would you build a system that requires blocks of keys to be contiguous?
this did actually happen. I rebuilt it becuase it was my job. I didn't specify it, needless...
Tim Wilkinson
"If it doesn't work in practice, you're using the wrong theory"
- Immanuel Kant
June 19, 2006 at 8:16 am
If by 'new to the programming world' you mean new to SQL or even programming in general, you are completely ****ed unless by some miracle the untested rough attempt below...
Tim Wilkinson
"If it doesn't work in practice, you're using the wrong theory"
- Immanuel Kant
June 19, 2006 at 7:34 am
closest is:
exec
sp_help N'EMP'
or select from a system view/table (depends on SQL version) with N Tim Wilkinson "If it doesn't work in practice, you're using the wrong theory"
- Immanuel Kant
June 19, 2006 at 5:24 am
nileshsane: Every SQL statement uses an isolation level. Maybe Jeff means: only reports should use READ UNCOMMITTED since they are generally running in a read-only environment?
BTW, using NOLOCK means you...
Tim Wilkinson
"If it doesn't work in practice, you're using the wrong theory"
- Immanuel Kant
June 19, 2006 at 5:16 am
Viewing 15 posts - 256 through 270 (of 582 total)