On 21/12/2008, at 3:25 PM, Paul Davis wrote:
Hmm. I don't think so. Technically we could change the implementation
to feed reduce functions a single map-row at a time which would be
O(N). Might end up violating some (incorrect) expectations of reduce
input, but I'm pretty certain it shows that it's O(N).
But you have to combine it with previously reduced input, which means
fetching a number of previous results, with a cost that isn't clearly
related to N. Isn't the number of reductions needed to be combined is
going to be related to key distribution?
Antony Blakey
-------------
CTO, Linkuistics Pty Ltd
Ph: 0438 840 787
It is no measure of health to be well adjusted to a profoundly sick
society.
-- Jiddu Krishnamurti
