[ Proposal ] [Draft] Variable Bounding and UnBounding Time

Cosmos Hub Proposal Draft on Variable Bounding and UnBounding Time based on the command and suggestions of the previously discussed pre-draft.

Google Doc with enabled comment option:


  • Background and Statistics
  • Definition of the Centralization Risk
  • Definition of Increased Bonding and Unbonding Time Funciton
  • Introduction of Enlightened Democracy Concept as approach to decentralization
  • Enable Variable Bonding Time and minimum Bonding Time
  • Enable Variable, Increased Unbonding Time
  • Enable instant undelegation/cancelation during bonding time

What does mean BPoA ?

Bonded Proof of Authority

I had a look at your proposal but I still do not get why enabling a variable bonding/unbonding time would be a better solution than enabling a variable minimum commission fee (see my comment here)

If I take myself as an example, I picked my validators depending essentially on the commission fee they propose ! I selected 3 validators with 0% commission fee and 3 others with commission fee between 2 and 5%, but this is just for the sake of it. If I felt the need to redelegate now, I would probably pick only 0% commission fee validators. And I do not particularly care about unbonding time.

So according to me, setting a variable minimum commission fee that depends on validator current voting power would be the most efficient way to incite users to delegate or redelegate to smaller validators. Also this rule appears simpler to me and does not appear more restrictive than setting a variable unbonding time. Or did I miss something ? Is there any other pros with your approach ? I admit I certainly did not study this question as much as you, so I am genuilely curious to understand why it would be better.

Because it is impossible to enforce commission fees, this proposal argues that if there is 1 rule there will be 2 goarounds. Instead of making something exploitable like minimum fees and falling into the trap that all governances in the world fall into - add bureaucracy like rules that then require even more bureaucracy and so on we should rather create a protocol level code of conduct that raises people awareness and combat centralization via education - Enlightened Democracy. I also proposed to crate and add to explorers a “centralization risk factor” which will raise awareness about the issue.

You mean it would be impossible to technically and automatically enforce a variable commission fee ? If it is the case then indeed this would not work for sure.

Yes it is technically impossible to enforce commission fees and on top of this it will lead to fraud. In less then 1 day after minimum commission fees are introduced we would see 50% of validators changing monikers to “Bla Ble Validtor | 0% commissions” and simple scripts returning money back to the users.

Then not so long later fraud through obscuring would occur where people create confusing rules regarding when they actually send the money back and so on, potentially causing huge outrage and backfire in the community.

So not only it would not work it would likely harm the network in the public eyes.

The complementary solution to my proposal is probably minimum guaranteed income to validators below certain threshold and rank allowing small fishes to set fees to 0 as they bootstrap. This is also large topic how to do it right and requires second paper to spec.

Regarding this I think that an additional commission for Validator based on the size of your validator greatly encourages decentralization, but is that if we add a variable in the time of unbonded period also in proportion to the size of the validator we get that decentralization is accelerated by reducing the time of linkage to minor validators

Not sure if they are complementary but definitely those foreign proposals are totally independent and should not be road blocking each other regardless which ones are pushed first.

Complementary Article to the topic, touching the philosophy behind proposal:

In my opinion, a period of variable unbonding must be accompanied by another variable that forces validators to loose her delegations exceded, otherwise they prefer to keep their delegations so as not to lose their position in the current Validator, thus causing an effect contrary to what was expected.

Could you define desdelegation a little more, what does this means.

1 Like