Eblob got Snappy compression support

Kudos to Google for opensourcing it.

I did not yet test its performance, but it compresses our small common XML about 2 times.
API was extended to support per-command flag to compress data or not. Eblob stores this information in its structure, so reading detects this and automatically decompresses data.

So, to store textual data elliptics and eblob look very promising, getting that Cassandra and MongoDB do not yet support compression (although there are tickets back from 2009) :)