RT @Kasabi: There are recipes to work with our datasets and Google Refine, Heroku, Python and others: http://t.co/yOFe2eoQ. What else should we create?
RT @cowtowncoder: One more update for [http://t.co/0KcnZgbo], Java compressor benchmark, with LZF 0.9 -- Snappy still fastest (+10-15%), LZF more compact +4%