Distributed consensus revised
blog.acolyer.org
blog.acolyer.org
I really like how understandable it is.
The needs for consensus with byzantine fault tolerance for crypto systems and the needs of consensus for distributed systems via paxosy or rafty algorithms are miles apart. The fundamental mathy research is related but the communities and the algorithms for each are quite different.
https://arxiv.org/abs/1711.06030
(Published version, closed access: https://dl.acm.org/citation.cfm?id=3278739)
Aziz, Haris; Lee, Barton E. – Sub-committee Approval Voting and Generalised Justified Representation Axioms (2017/2018)
Abstract:
"Social choice is replete with various settings including single-winner voting, multi-winner voting, probabilistic voting, multiple referenda, and public decision making. We study a general model of social choice called sub-committee voting (SCV) that simultaneously generalizes these settings. We then focus on sub-committee voting with approvals and propose extensions of the justified representation axioms that have been considered for proportional representation in approval-based committee voting. We study the properties and relations of these axioms. For each of the axioms, we analyze whether a representative committee exists and also examine the complexity of computing and verifying such a committee."
Citations: https://scholar.google.com/scholar?cites=5804298394619698922
And then, perhaps, that theorem of infamity by that dude who might have gone by—some name(, like, uh, )—'Aumann'
Consensus in voting systems is concerned with selecting either a single winner or a ranked order from a set of candidates given the diverse preferences from a pool of voters. It concerns the protocol by which they communicate their preferences (vote favorite, pairwise, ranked preferences, etc.) and the protocol by which they're combined to pick a winner (consensus) from the population. The focus is on minimizing how objectionable the winning candidate is over all voters.
Distributed consensus is more concerned with agreement of a proposition between set of identical nodes (i.e. no preferences) that communicate by sending messages over an asynchronous, lossy, reorderable, channels. The focus here is about mitigating weakness of the communications channels and provide stronger guarantees beyond that.
This is precisly the problem with electronic voting. It is not about creating a process that is mathematically sound — it is about creating a process the average voting helper can transparently follow, verify and trust in. Every electronic system, no matter how transparent is always a black box to everbody who didn’t either design it, or has both the skill and time to take it apart.
A mathematically sound result that average people can doubt, because it is a blackbox to them, is useless for any meaningful vote.