On Tue, 31 Jul 2001 ghaverla@freenet.edmonton.ab.ca wrote:
>
> I'm sure you can do a search for a RNG without replacement as
> easily as I can. Yes, a person can always keep track of what
> numbers have already been picked and reject duplicate picks.
> If the number of picks you are making is much smaller than the
> space you are picking from, this would work fine. Especially
> using a smart (not linear) search for the possible duplicate.
> If the number of picks is getting close to the size of the
> stack, rejection is going to be a big loser. Any program
> that draws cards from a deck is going to use this type of
> without replacement method, as some card games deal the entire
> deck.
I don't think it is wise to do it externally if you can let the db do it
using unique columns. The method you suggested just sounded new to me,
I would have preferred a hash table or a tree structure.
- Einar Karttunen