I have actually considered doing exactly this, however there are currently 15 or so other procs that actually insert into ContactDetails table, so I'd rather not have to go through...

I can't add an identity column on the table as there's already one there... however I've tested this locally and it's significantly faster than I expected... I think I have...

Remi:  I CAN add columns, but that'll definately slow stuff down, no?

Farrell:  if I remember, in 6.5 you had to explicitly call indexes into use, but 7.0 and 2000 do...

I'm guessing the syntax for this would be something like

declare @Id int

select @Id = min(Id) from ContactDetails

while @Id < (select max(Id) from ContactDetails)

begin

Update cd

set CurrentValue = 'N'

from ContactDetails cd

where Id not in (select max(id)...

do you have any syntax that I could see regarding this?  I've never had to force the use of an index...

nope, no triggers... tried that, was REALLY slow

there are a ton (7 - 8) indexes on it... but these should only speed it up.

I have actually tried this using a couple different methods, all slow down the insert rather significantly... inserts of 7500 users goes from 45 - 60 seconds to 15 - 20 minutes, not good...

just your suggestions...

30 second run is definately better than a 8 - 9 minute run

thanks for the input... now time to play with...

WELL said

There's one thing everyone here has missed so far... security.

When we allow developers to write the code they know straight into the db, we pretty much immediately remove sql injection...

ok, so playing with the function idea.  There's just 1 problem.  the proc that I was trying to run in my original update statement dealt with it's own temp tables,...

thank you, and agreed, I'd love to be able to use more static sql...

however, I have to create a table with a n number of columns, and it's only applicable to...

oh joy... can't wait to do my move at the end of next month...

luckily for me, I will also be moving the 400lb dell rack with my 3 servers, kvm,...