Hello again.
A lot has happened since I wrote this post, and much of it has better informed my view of the responsibility of the oDAO and its members’ roles within the ecosystem. I figured I’d post a quick update on what my priorities are for the oDAO.
Note that I am a member of the Rocket Scientists, who have an oDAO seat operated by Ken, but am still interested in operating an oDAO node myself.
- First, to improve our ability to detect bugs, I’d like to build a custom watchtower implementation, with its own suite of tests. This would probably require a formal specification of watchtower duties, similar to Joe’s treegen specification, to truly be a “clean room” alternative.
- Second, I’d like to implement treegen separately. In the interest of breaking up the work, custom watchtower would likely use stock treegen until stable.
- Once stable, I’d like to actually separate the treegen image from the watchtower image- oDAO members should be able to pick which watchtower and which treegen they would like to use independently
Unfortunately, two watchtowers is not much better than one watchtower in terms of consensus, though it does still help us identify bugs with either implementation. So once stable, I’d like to support the development of a third implementation of each of watchtower/treegen by a third party, and would work to encourage oDAO members to coordinate and ensure that no implementation accounts for 50%+ of the vote. This long-term goal would improve the consensus model for the oDAO, as well as the security (a supply chain attack would now have to successfully compromise a majority of software sources, instead of just one).
This work can be carried on to a larger conclusion- the implementation of an alternative smartnode stack- which would provide additional security to the pDAO as well.
My motivations here come from a perception that the team wants a robust oDAO, but isn’t positioned to improve these things themselves, as a second implementation of watchtower from the team itself would have diminished returns in terms of security and decentralization. Instead, the pDAO and oDAO should take the reins on these projects, so I have prioritized this work to start shortly after my responsibilities to the Rescue Node quiet down.