ansnunez

Avoid using conditional variable to manage GETSTARTPOWFROMSEED requests

1

Consider using a buffer instead, since we can't anticipate the number of requests a lookup will receive

Re-evaluate handling of null pointers

1

After code review on #566, specifically on handling raw pointers returned by a third-party library in that PR, a convention for uniformly handling pointers is needed. For example, we have...

Protobuf improvements

1

## Variable naming Follow-up todo based on #590 review. Follow the protobuf style guide [here](https://developers.google.com/protocol-buffers/docs/style). If the field name in `.proto` file could be `this_is_field_name`, then the generated function name...

Enforce getting the return value of a function

2

We had a bug last week of this nature: ``` int funcSetXToZero() { return 0; } int main() { int x = 1; funcSetXToZero(); // failed to set x return...

Avoid using blocknum and epoch num interchangeably

1

In #638, the block number parameter is actually set to epoch number in all consensus types. This seems quite confusing, and as discussed in the PR it is also prevalent...

Missing ShardCommitFailureHandlerFunc for some consensuses

2

For some consensuses, `NodeCommitFailureHandlerFunc` is defined but `ShardCommitFailureHandlerFunc` is not. Need to define the logic for how commit failure for the shard should be handled in these cases.

Create more exhaustive tests for block serialization

1

Check completeness of unit tests for serialization and deserialization of blocks. It is tested implictly when testing persistent storage.

Consensus ID is not set correctly

1

For DSBlock consensus, `consensus_id` is fixed to 0, but for FinalBlock consensus, we are using `m_mediator.m_consensusID`. Other consensus types should be checked as well.

Graceful termination

2

Implement some signal handler that, once triggered, does cleanup and termination.