I don't understand how a row by row operation can ever be more efficient than a set based operation, let alone by a factor of 100.
 
As far as sql vs oracle, microsoft made huge ground on oracle with sql 2000.
"Mark Blair" <[EMAIL PROTECTED]> wrote in message news:22362@cfaussie...
Cursors aren't bad always, cursors _can_ increase response times by a
factor of 100 (give or take 100%) on SQL Server, this is more generally
an issue when dealing with large resultsets (100,000 plus) ... it should
also be noted that certain databases such as Oracle have little
noticeable performance impact when dealing with cursors (not trying to
start an Oracle versus SQL Server thread, I know which one I prefer
given dollars are not a concern)

cheers
Mark
 
>Cursors are bad mmmkay.  There is pretty much no excuse for ever using
>cursors in a stored procedure, there is nearly always a better way of
>writing the query. You do realise when you use a cursor every iteration is
>like doing another query to the database, i can't begin to tell you how
>inefficient they are.

>But even better yet used derived tables. (see
>http://www.sql-server-performance.com/jg_derived_tables.asp)

>http://www.sql-server-performance.com/transact_sql.asp
---
You are currently subscribed to cfaussie as: [email protected]
To unsubscribe send a blank email to [EMAIL PROTECTED]

MX Downunder AsiaPac DevCon - http://mxdu.com/

Reply via email to