This looks amazing! Hadley, Wes, what are your thoughts on how to implement compression? I recall some open source columnar datastores (e.g. infobright) that

This looks amazing! Hadley, Wes, what are your thoughts on how to implement compression? I recall some open source columnar datastores (e.g. infobright) that achieved very VERY fast compression rates with just a few tricks: https://news.ycombinator.com/item?id=8354416 https://news.ycombinator.com/item?id=8354416 In particular, compression is extremely fast for columnar datastores (its the same type one after the other). Since a lot of times the data is sorted by some ID (date, individual, etc.), you should see large improvements in both speed and disk space.