Bisq DAO Cycle 20

The figures on this page are compiled from data files generated by the Bisq software. You can verify everything yourself by running these scripts on GitHub.

Cycle Started9 Dec 2020 / Block 660667

Cycle Ended9 Jan 2021 / Block 665346

Supply Change— 87,828

Governance0 of 0 proposals accepted

Network Funds Transfers

BSQ Amount # of transactions
 Proof-of-burn¹ 58,000 1
 Trading fees² 29,746 3,177
 Compensation request fees 44 22
 Blind vote fees 26 13
 Reimbursement request fees 8 4
 Proposal fees 4 2
Total Burn 87,828 3,429
 Compensation³ 0
 Reimbursements⁴ 0
Total Issuance 0
Net BSQ Supply Change⁵ 87,828

¹ Proof-of-burn includes trading fees paid in BTC and disputed BTC deposits for trades that went to arbitration (see docs for more details). Funds may be accrued and burned in different cycles, so proof-of-burn figures do not map directly to activity in their cycles.

² BSQ trading fees only. BTC trading fees are included in proof-of-burn.

³ See more details on GitHub.

⁴ Over time, the net impact of reimbursement issuances on BSQ supply is close to zero, as corresponding amounts of BTC are burned through proof-of-burn (see docs for more details).

⁵ Decreases in BSQ supply are good.

Critical Cycle Notes

Results of Cycle 20 of the Bisq DAO were invalidated because voting nodes could not establish the required 80% supermajority consensus on the prevailing data view, ostensibly because of Tor issues that surfaced in the last few blocks of the cycle.

All proposals failed to be approved or rejected. No BSQ was issued.

Stakeholders will make new proposals in Cycle 21 for both Cycle 20 and Cycle 21. The BSQ rate will be 0.75, which is halfway between the rates originally approved for Cycle 20 and Cycle 21 (see discussion).

In the vote reveal phase (which lasts 450 blocks), each Bisq node publishes a Bitcoin transaction that includes a hash of an ordered list of all voting data objects it received during the blind vote phase. This is done to establish a prevailing view of voting data—since the network is eventually consistent, each node cannot be trusted to have the exact same set of data, so each node must compare the set of data it received with the set of data every other node received.

Voting weight is the metric used to resolve differences. Recall that each ballot submitted during the blind voting phase has a BSQ stake attached to it—this same voting stake is used to weight voting data views in the vote reveal phase. Nodes with a collective minimum of 80% of the cycle’s voting stake must agree on 1 data view, otherwise, the whole voting cycle is invalidated.

Tor network issues caused instability on the Bisq network for several days starting toward the end of Cycle 20’s vote reveal phase.

As a result, one or more large stakeholders seemed to miss some ballots, which caused the prevailing data view consensus to dip below 80%:

}, HashWithStake{
}, HashWithStake{
}, HashWithStake{

Observe the four HashWithStake items above. The first is the majority data view with a ~72% stake. The next three are minority stake views that are probably missing ballots. If data had propagated correctly to that second node (or group of nodes), the 80% threshold would have been cleared. But it didn’t, so the cycle was invalidated.

It is important to note that this outcome is by design and intentional. In addition to being a measure to handle network instability, the 80% threshold is meant to serve as a safegaurd against attackers—someone would need more than 80% of the stake in a cycle to unilaterally influence it.

< See more DAO data on the Dashboard