At 12:31 PM 2005-08-20, you wrote: >Assume your update is something like this: > >UPDATE table >SET rank = rank + 1 >WHERE rank >= 33 > >The update will be really fast, even with a large number of records. Since >this is a web application, I'm going to jump out on a limb and assume that >SELECTs far out number the UPDATEs and INSERTs. If this is true, then even >if this WAS somewhat slow it wouldn't be a big deal in the big picture. Yes, this is the easy part, but if we already have have a number 34, this operation would leave us with TWO records of rank #34, which is what I'm trying to avoid. I was also hoping to avoid looping over each record (from 33 to the end). I'm not sure that I can avoid the overall update to all the records. My concern was that I might end up with 500+ records, and that it would consume lots of resources, and that it increases the risk of corruption to the database. > > Seems to me that updating 64 records with a different > > rank might be rather inefficient, especially if we have a > > lot of records. > >BTW, 64 records is always 64 records, even IF you have lots of records. ;-) <chuckle> Got me. Frank Marion lists at frankmarion.com Keep the signal high.