Re: Selecting K random rows - efficiently!

> How important is true randomness?

The goal is an even distribution but currently I have not seen any way
to produce any kind of random sampling efficiently. Notice the word
"efficiently". The naive way of taking a random sample of size K:
    (SELECT * FROM mydata ORDER BY random() LIMIT <K>)
is clearly not an option for performance reasons. It shouldn't be
necessary to explain why. :-)

> Search the archives, there have been solutions proposed before, though
> they probably arn't very quick...

As the subject suggests, performance really matters and searching the
archives only results in poor solutions (my first post explains why).