Polkadot Governance v2.
What we have learned from Polkadot Decoded 2022.
Polkadot Decoded 2022 was this year's largest Polkadot event, with more than 100 talks.The keynotes for the two days were given by Dr. Gavin Wood, and the first one was on the changes in the Polkadot governance mechanism. So what's wrong with Polkadot governance now?
Currently Governance 1 is Tricameral:
- Technical Committee (consisting of 3 members) for fast-tracking
- Council (13 elected members) for executive matters
- Referenda for Root-level stuff (Adaptive-quorum basing, conviction-voting, delegation)
The positive sides are:
- Conviction-voting (weight sentiment toward long-term participants)
- Delegation (allow holders to participate without direct voting)
The negative sides are:
- Slowness (28 + 28 days voting and enactment for every referendum)
- Clumsiness (only one referendum running at once, alternating council & public)
- Inaccessibility (low chance that proposal will ever reach voting)
The Solution: Governance v2
LOWER BARRIERS (Remove the Council and Technical Committee + Unlimited referenda on everything all at once)
- For proposals to happen
- For referenda to get opened
- For people to vote
How to make Unlimited Referenda safe?
- Limit number of active decisions
- Restrict possible repercussions
- Require high turnout/approval
- Increase deliberation/reacting periods: before voting starts, during voting “confirmation”, after voting ends
Gov1 used restricted activity & trust: safe but slow, clunky and overly centralized.
Gov2 uses many more elements to ensure optimum & adaptive balance between safety and agility.
Restrictions of throughput and time depend on theoretical possible damage by the referenda.
Classifying possible repercussions.
Not all proposals are equal.
If we can classify a referendum as as being like low impact then we don’t need to be quite as conservative with our safeguards.
Proposal is two things: Operation (what should happen) and
Origin (who/what authorizes it). Most Operations require a particular Origin. Different Origins have different levels of privileges.
There are many kinds of Origins, and their impact may vary from low to high. There is a different Track (pipeline) for each Origin type. Each of these tracks can have different parameters. Low-impact origins need less stringent safety requirements than high-impact origins since they are less dangerous.
New simple Life Cycle of a Referendum:
- Proposing. Anyone can propose a referendum at any time, no limit to number proposed & small deposit for each, also everyone can vote on it.
- Deciding. A second, “deciding” deposit must be paid and the track’s lead-in period has to pass. If fully engaged, eligible referenda with the most aye votes are prioritized once there is track capacity.
- Confirming. To be approved, a referendum must be passing continuously for the confirmation period. Controversial referenda may cycle in and out of passing/confirming a lot while deciding.
- Ending. If the deadline of 28 days elapses without confirmation, the proposal is rejected. Passing throughout the confirmation period means that proposal is approved.
- Execution. After approval, the operation is scheduled for execution with the proposed origin. A minimum enactment delay will be adopted as a safety measure.
Updated Voting Mechanism:
- The number of aye votes as a proportion of the maximum possible number of votes.
- It is measured ignoring conviction as well as taking it into account.
- Nay votes are not counted, otherwise it could be counter-productive to vote at all.
- Thresholds change over time: required % of support and approval start conservatively high. Levels reduce as the vote proceeds, ending at the lowest possible point.
Other features (under development)
Multirole Delegation as the new feature of Gov2.5: community members will be able to delegate to a different voter for each track (one delegation for tips, another for upgrades e.t.c.)
“This is a way of saying I’m okay with this account using the voting power from my account and you can do that without paying any transaction fees.”
In addition to it there is free undelegation what this means in combination is that wallets when they build these features in can have a very low friction interface that wouldn’t require the user to spend any funds.
How do we ensure that a proposal is safe? Three things that we need to do:
1. Track everyone’s level of expertise;
2. Allow everyone everyone to register their sentiment;
3. Aggregate it all
What is Polkadot Fellowship?
Is not a body that can do anything directly to the Polkadot protocol, to any of the protocols, to any of the networks. It has no power to change the protocol, the only thing it can do is declare that it believes a particular proposal is safe and time critical.It’s an oracle nothing more.
Here is nice article from Coindesk By Ian Allison