Postmortem - Bringing mesa.eth back online

Hey everyone,

First I want to congratulate the collective on showcasing the ability to reach absolute majority in less than 24 hours. This demonstrates the ability of the DXdao to pass proposals without relying on boosting. It is also proof that the members care and are active participants in governance.

Unfortunately, our attempt to bring Mesa.eth online with an absolute majority failed because of a technical reason which I want to explain here:

On the 7th of May a Dxdao member made a proposal to change mesa.eth´s public resolver contract from the newest version to an older version which was not configured with any IPFS content and without creating a daotalk.org thread for community discussion.

The main goal of this proposal was to enable the collective to update mesa.eth through the connection to the older version of the public resolver while the proposal to add the connection to the newest version of the public resolver was still pending.The proposal to configure mesa.eth to the old public resolver passed on the 20th of May and since then mesa.eth is not resolving to any IPFS content because the content hash isn`t configured.

The downtime of mesa.eth was quickly detected and a few members created a proposal to set the IPFS content hash to bring mesa back online as soon as possible via an absolute majority consensus vote.

The absolute majority was reached in less than 24 hours and the vote passed successfully. Unfortunately, this proposal did not set the IPFS content because the older version of the resolver has a reference to the old ENS contract which is not aware of the owner/resolver of mesa.eth because it was registered on the new ENS contract.

What the collective can learn from this:

  • If a proposal changes the on-chain system of the Dxdao, create a Daotalk thread first and allow at least 3 days to let the community give input on the requested changes.

  • The collective should formalize a process of how changes on the on-chain system should be done in a secure, thoughtful and coordinated manner.

  • Once the Dxdao on-chain system is interacting with other contracts like the ENS contracts, it is necessary to get a complete overview and understanding of how those third-party protocols work. Which should result in a guidebook specifically for the ENS contracts(and any other protocol the DXdao will connect to).

  • The ENS migration added complexities which were not fully appreciated and accounted for by the developers involved in these updates. Going forward we must make sure the community has a clear understanding of the systems with which it is integrated.

9 Likes

Learning from mistakes, things happen. Thanks for the explanation.

4 Likes