mlw <[EMAIL PROTECTED]> writes:
> My problem is, if you do a select by the user name, it does an index
> scan. If you do a select from the whole table, ordered by the user
> name, it does a sequential scan not an index scan. It is arguable that
> this may be a faster query, but at the cost of many
I got an interesting question, and I can probably see both sides of any debate,
but.
Say you have a fairly large table, several million records. In this table you
have a key that has a fairly good number of duplicate rows. It is a users
favorites table, each user will have a number of entries
