instead of triggers were introduced in sql 2k. What's the problem with the code you created?

I never had to either... but I can see the need. Imagine that the server can't keep up with the demand and has to queue to inserts of the...

I didn't check the link, but this might give you what you need :

Create table...

GO

exec sp_addextendedproperty N'MS_Description', N'This table must contain ONLY 1 LINE or Facturation.adp.Km (form) will start...

Yes I speak french. Also if the method Ray posted doesn't work we'll need some sample data you need to generate.

Here's the rest of mine if anyone's interested :

CK : check constraint

DF : default constraint

FK : Foreign key --for objects and columns

PK : Primary key --for objects and columns

fn...

Usually for sampling. Something like an election where you want to get an estimate every X minute without recompiling everything. But there are always other methods to do...

That's where you see who's the boss on the server .

ok, ok, I'll give you my vote if you insist.. Send the t-shirt at the same adress as all the books I ordered .

Select * from Information_schema.columns where table_name = 'TableName'

Search the script section on this site... you'll find what you need.

Ya, 33k records can cause a lot of page reads if it's fragmented... but not as much as an useless scan .

... and everyone else who followed the same course as I did so that would be a few hundreds .

HTH.. Please post back your findings... I want to see how this turns out .

CREATE TABLE [ObjADPSUsage] (

[PkObjADPUsage] [int] IDENTITY (1, 1) NOT NULL ,

[FkObjADP] [int] NOT NULL ,

[DateUsage] [datetime] NOT NULL CONSTRAINT [DF_ObjADPSUsage_DateUsage] DEFAULT (getdate()),

[FkUser] [int]...

I already posted something like that. Look carefully at the where in the clustered scan. You'll most likely find the condition to be something like this : where...