Nate Rush
Nate Rush
# Issue NOTE: This just a thought I'm putting here so I don't forget it. It probably makes no sense Currently, messages do not arrive in order; because messages include...
# Issue Checking if one block is "in the blockchain" (in the prev block pointer chain) of another is linear in the length of the chain. Ideally, we could check...
# Issue Currently, we define [some constants](https://github.com/ethereum/cbc-casper/blob/master/simulations/network_delay.py#L4) that the network delay functions use to calculate the delay on certain messages. These constants will also include the [peer connectivity matrix](https://github.com/ethereum/cbc-casper/issues/155), when...
# Issue Currently, some protocols allow validators to make (random) decisions when creating a new estimate. For example, see [here](https://github.com/ethereum/cbc-casper/blob/master/casper/protocols/concurrent/concurrent_view.py#L31) or [here](https://github.com/ethereum/cbc-casper/blob/master/casper/protocols/sharding/sharding_view.py#L65). This limits how well we can test these...
# Issue Currently, the sharded blockchain and concurrent schedule protocols do not have safety detection implemented. # Proposed Implementation - Define when two messages conflict. [There's a placeholder here](https://github.com/ethereum/cbc-casper/blob/master/casper/protocols/sharding/block.py#L46) in...
# Issue We need more data structures to come to consensus on! Let's add consensus for a commutative data structure! If you want to read more about CRDTs, see a...
# Issue Currently, we don't check message validity. To see if a message is valid, its estimate should be justified. For some protocols, checking that an estimate is valid requires...
# Issue - What's wrong? - How are things currently? - How should things be? # Proposed Implementation Fill this out as best you can, and we'll start the discussion...
# Issue Currently, we have no way of figuring out how efficient different safety oracles are in different situations. It would be nice to be able to compare these different...
# Issue Currently, we start with consensus on a single validator set. Instead, we'd like to be able to change the validator set while the protocol is executing. # Proposed...