Forking as a last-ditch governance mechanism

[Arachnid]

Forking is a great mechanism in applications like this to provide people with a way to opt-out from a decision they really don't like.

One possible way to do this: Let people vote "yes", "no" and "over my dead body". If a proposal is accepted, anyone who voted "over my dead body" gets transferred with their assets to a duplicate version of the contract.

[decanus]

We did document an exit mechanism, however I guess a forking mechanism would be interesting to implement to allow multiple parties to create a new DAO in case they do not align with the current DAO. Ideally a forking function would make the process easier, for example redeploying, moving funds etc. Important to note here are the security considerations.

[decanus]

@Arachnid Do we really want to automatically group the "over my dead body" group into the same duplicate contract, would it not be smarter to allow a manual forking process where people can vote on a split, those voting for "yes" get moved along with their funds?

[Arachnid]

@decanus ISTM all the "over my dead body" people want the same result, but I could be mistaken. Can you give an example where two people voting that would want to end up in different forks?

[decanus]

@Arachnid just because we agree on some things doesn't mean we agree on all things. However I guess with the "over my dead body" system we would still at some point reach optimal groups if there are enough forks, however with people deciding to join certain forks based on philosophies or whatever this process may be sped up.

[lkngtn]

I think the more important consideration for forking is not wether you redeploy contracts but how you divide the DAOs assets. Assuming all assets are divisible they can be split between the two DAOs proportional to the split, however, it may make more sense to simply give users the option of exiting with a pro-rata share of the assets in response to a decision they absolutely disagree with. Then if they want redeploy the contracts and start a fork of the DAO they can.

Using that approach also simplifies some of the other messy details related to non-forkable assets that a dao would likely be associated with such as brand, domains, social accounts, etc.

[decanus]

@lkngtn I guess the question we need to answer here is how often we can expect a fund to fork. In a scenario (which I would endorse) where we see forks occurring at every proposal @Arachnid's system may be more beneficial, however I feel this can also be solved through a system like mine with a seperate fork proposal. There may be a certain scenario where the forks occuring are so high that it is more efficient to have native forking, however that mechanism looks.

[decanus]

Also another question is, in a governance system do we want to motivate people to fork, or motivate them to remain together. I like the idea of motivating people to fork, I would like to see the environment this can create.

[decanus]

@decanus:

The forking stuff raises an interesting question how many forks are sustainable, and when do we hit the threshold which shows us that our governance is ineffective due to people constantly forking

@lkngtn:

Alternatively what is the size of an organization that is optimal