RE: One-Block Irreversibility for Delegated Proof-Of-Stake (DPOS)

[marki99]

I think it's a great idea.Two questions:

1-Isn't OBI basically what tendermint does?

2-Doesn't OBI make HAF redundant, since the whole point was to enable devs not to deal with fork handling?

Comments

[blocktrades]

1-Isn't OBI basically what tendermint does?

Don't know, haven't read about tendermint, although heard of it in passing. But several of the implementation details of OBI are pretty specific to the way Hive does DPOS, so to be the same would require very similar DPOS functionality. But some commonality wouldn't be surprising in any case.

2-Doesn't OBI make HAF redundant, since the whole point was to enable devs not to deal with fork handling?

No, but it makes HAF less critical. Why HAF is still important:

• Forks can still happen, even with OBI. OBI just speeds up irreversibility in the normal case of a well-connected network.
• HAF isn't just about fork handling, it's also about scalability (very high API performance without loading down nodes) and easing development (customizable API, use of standardized languages for development, etc).