Click here to monitor SSC
SQLServerCentral is supported by Red Gate Software Ltd.
 
Log in  ::  Register  ::  Not logged in
 
 
 
        
Home       Members    Calendar    Who's On


Add to briefcase ««12

Shift week to Wed - Tues Expand / Collapse
Author
Message
Posted Friday, September 27, 2013 6:23 AM
SSC-Enthusiastic

SSC-EnthusiasticSSC-EnthusiasticSSC-EnthusiasticSSC-EnthusiasticSSC-EnthusiasticSSC-EnthusiasticSSC-EnthusiasticSSC-Enthusiastic

Group: General Forum Members
Last Login: Yesterday @ 12:29 PM
Points: 127, Visits: 239
paul.j.kemna (9/27/2013)
Isn't SET DATEFIRST a sever wide setting?


Nope. Only affects the session. I've been developing code to create and update a master calendar table for our organization so I've been hip-deep in dates this past week.


____________
Just my $0.02 from over here in the cheap seats of the peanut gallery - please adjust for inflation and/or your local currency.
Post #1499344
Posted Friday, September 27, 2013 8:06 AM
Hall of Fame

Hall of FameHall of FameHall of FameHall of FameHall of FameHall of FameHall of FameHall of FameHall of Fame

Group: General Forum Members
Last Login: Yesterday @ 6:25 PM
Points: 3,121, Visits: 11,397
Jeff Moden (9/26/2013)
Michael Valentine Jones (9/26/2013)
Jeff Moden (9/26/2013)
Michael Valentine Jones (9/25/2013)
That's not really the best way to go because it depends on the setting of a run-time session parameter, DATEFIRST.


Although I certainly prefer you method, I have to ask... why does everyone have such a problem with that? It's no different than setting a variable that would be used as a constant. The only time that such a setting change would be a problem is if it were in a function, which can't be done TTBOMK.


I wanted the OP to be aware of the limitations of their method. They are certainly free to ignore my wonderful advice.

I prefer to have code that works as expected under the widest possible conditions, especially if it's no harder to code.

People may be operating with different default languages than our "normal" US English, and that can change the setting of DATEFIRST, and the client application can also define that setting in an unexpected way.


Absolutely agreed across all points especially the part about it not being any more difficult. I was actually talking about putting SET DATEFIRST into the actual code, though. If someone can't figure out the integer math, I can't see why you couldn't use SET DATEFIRST for the session. People seem almost phobic about its use.


If you set DATEFIRST to something other than the expected value, there could be downstream issues in the next procedure or code that gets executed if it depends on the setting of DATEFIRST. Or if your code depends on the setting of DATEFIRST, you don't specifically set DATEFIRST, and upstream code has modified DATEFIRST to an unexpected value.

If you are going to use DATEFIRST, it is probably best to set it specifically and then reset it to the default if you are setting it to a non-default value.

If you are using connection pooling, I believe that the value of DATEFIRST does get reset to the default by the API stored procedure sp_reset_connection (unlike the isolation level )

set nocount on
set datefirst 7 -- Normal US English setting
select DW1 = datepart(dw,'20130927')
go
set datefirst 3 -- Set to Wednesday
select DW2 = datepart(dw,'20130927')
go
exec ('select DW3 = datepart(dw,''20130927'')')

Results:
DW1
-----------
6

DW2
-----------
3

DW3
-----------
3


Post #1499401
Posted Friday, September 27, 2013 9:37 AM
SSCommitted

SSCommittedSSCommittedSSCommittedSSCommittedSSCommittedSSCommittedSSCommittedSSCommitted

Group: General Forum Members
Last Login: Today @ 8:25 AM
Points: 1,963, Visits: 2,899
Why mess with the DATEFIRST setting when it's not necessary? There are methods which can do the calculation simply without needing a specific datefirst setting, so why hassle with it and take chances?

SQL DBA,SQL Server MVP('07, '08, '09)
"In America, every man is innocent until proven broke!" Brant Parker
Post #1499472
Posted Friday, September 27, 2013 3:58 PM


SSC-Dedicated

SSC-DedicatedSSC-DedicatedSSC-DedicatedSSC-DedicatedSSC-DedicatedSSC-DedicatedSSC-DedicatedSSC-DedicatedSSC-DedicatedSSC-DedicatedSSC-DedicatedSSC-DedicatedSSC-Dedicated

Group: General Forum Members
Last Login: Today @ 12:14 AM
Points: 36,735, Visits: 31,186
ScottPletcher (9/27/2013)
Why mess with the DATEFIRST setting when it's not necessary? There are methods which can do the calculation simply without needing a specific datefirst setting, so why hassle with it and take chances?


Because, as strange as it sounds, some people just can't do the math on those simple methods.


--Jeff Moden
"RBAR is pronounced "ree-bar" and is a "Modenism" for "Row-By-Agonizing-Row".

First step towards the paradigm shift of writing Set Based code:
Stop thinking about what you want to do to a row... think, instead, of what you want to do to a column."

(play on words) "Just because you CAN do something in T-SQL, doesn't mean you SHOULDN'T." --22 Aug 2013

Helpful Links:
How to post code problems
How to post performance problems
Post #1499604
Posted Friday, September 27, 2013 3:59 PM


SSC-Dedicated

SSC-DedicatedSSC-DedicatedSSC-DedicatedSSC-DedicatedSSC-DedicatedSSC-DedicatedSSC-DedicatedSSC-DedicatedSSC-DedicatedSSC-DedicatedSSC-DedicatedSSC-DedicatedSSC-Dedicated

Group: General Forum Members
Last Login: Today @ 12:14 AM
Points: 36,735, Visits: 31,186
Michael Valentine Jones (9/27/2013)
Jeff Moden (9/26/2013)
Michael Valentine Jones (9/26/2013)
Jeff Moden (9/26/2013)
Michael Valentine Jones (9/25/2013)
That's not really the best way to go because it depends on the setting of a run-time session parameter, DATEFIRST.


Although I certainly prefer you method, I have to ask... why does everyone have such a problem with that? It's no different than setting a variable that would be used as a constant. The only time that such a setting change would be a problem is if it were in a function, which can't be done TTBOMK.


I wanted the OP to be aware of the limitations of their method. They are certainly free to ignore my wonderful advice.

I prefer to have code that works as expected under the widest possible conditions, especially if it's no harder to code.

People may be operating with different default languages than our "normal" US English, and that can change the setting of DATEFIRST, and the client application can also define that setting in an unexpected way.


Absolutely agreed across all points especially the part about it not being any more difficult. I was actually talking about putting SET DATEFIRST into the actual code, though. If someone can't figure out the integer math, I can't see why you couldn't use SET DATEFIRST for the session. People seem almost phobic about its use.


If you set DATEFIRST to something other than the expected value, there could be downstream issues in the next procedure or code that gets executed if it depends on the setting of DATEFIRST. Or if your code depends on the setting of DATEFIRST, you don't specifically set DATEFIRST, and upstream code has modified DATEFIRST to an unexpected value.

If you are going to use DATEFIRST, it is probably best to set it specifically and then reset it to the default if you are setting it to a non-default value.

If you are using connection pooling, I believe that the value of DATEFIRST does get reset to the default by the API stored procedure sp_reset_connection (unlike the isolation level )

set nocount on
set datefirst 7 -- Normal US English setting
select DW1 = datepart(dw,'20130927')
go
set datefirst 3 -- Set to Wednesday
select DW2 = datepart(dw,'20130927')
go
exec ('select DW3 = datepart(dw,''20130927'')')

Results:
DW1
-----------
6

DW2
-----------
3

DW3
-----------
3




Thanks, Michael. I keep forgetting about connection pooling. That's a REALLY good reason to not use it.


--Jeff Moden
"RBAR is pronounced "ree-bar" and is a "Modenism" for "Row-By-Agonizing-Row".

First step towards the paradigm shift of writing Set Based code:
Stop thinking about what you want to do to a row... think, instead, of what you want to do to a column."

(play on words) "Just because you CAN do something in T-SQL, doesn't mean you SHOULDN'T." --22 Aug 2013

Helpful Links:
How to post code problems
How to post performance problems
Post #1499605
« Prev Topic | Next Topic »

Add to briefcase ««12

Permissions Expand / Collapse