TODO

Plans for future:

- number tests
- PODify docs
- write up the specification of the proof file (and make sure it's OK)
- shorten Proof::new_from_share
- get rid of blessing hacks for new_from_binary, perhaps by making the classes
  inherit Data::ParseBinary::Struct and/or adding more adapters
- create a nicer framework for commandline tools
- make commandline tools get actual commandline arguments
- make the config file location more configurable
- add a proof splitter (to split a long proof into shorter ones)
- add a proof manager (that'll automatically find shares etc)
- add a proper proof creator
- use far_share_hash to jump by 100s to make proofs shorter

Later future:
- group up shares into merkle trees to reduce size of proofs
- create "authorities" that'll sign merkle roots of such trees to prove
  that shares actually existed at said times - this will allow more
  granularity in the timestamps
- create a p2p network where users can send data to be timestamped to miners,
  removing the need for everyone who wants to timestamp data to be a miner.
- optimize it, maybe even rewrite parts in C