Some thoughts about FOCIL:
I don’t want people to get the impression that FOCIL is somehow designed to push forward a scaling roadmap, it is not. The core proposal of FOCIL is to rely on multiple parties amongst the more decentralized validator set to preserve Ethereum’s censorship resistance properties by constraining sophisticated builders and force them to include transactions so they can’t arbitrarily censor.
Yes, the increased CR given by FOCIL can then be considered as a first step towards decoupling local block building from CR responsibilities, but we would also crucially need a good CR mechanism for blobs, and a way to be resilient against liveness failures if we were going towards this direction anyways. Whether FOCIL belongs in the light or medium node category (if that’s how we want to frame things in the first place, there are many possible variations) doesn’t matter much right now imo (as in, for the current proposal), because we would need to answer
h/t @Julian: Where exactly FOCIL is in the spectrum of 1-of-n or n/2-of-n doesn’t matter so much, we can adjust the mechanism either way (we can increase the number of includers per slot, or the size of ILs, etc)
A FOCIL node doesn’t need to execute transactions and also doesn’t need to hold all the state so it could potentially fit in the light node category as well. Ideally, we go towards a future in which home operators can participate in preserving Ethereum’s CR with very low hardware and bandwidth requirements, I personally think we can get there. Here’s a recent proposal in which nodes only have to store the minimal data required to preserve CR and maintain the public mempool called VOPS.
Genuinely curious to know more about what circles that have barely even started thinking about it? FOCIL isn’t really user facing, and we have talked a lot with core devs, researchers, builders, searchers, and MEV people. But if anyone reading this wants to learn more or has questions feel free to reach out anytime.
More generally, I want to make sure people realize that yes, FOCIL introduces a new role and a new primitive, specifically by imposing constraints on builders, and allowing multiple validators to have a say in what goes in Ethereum blocks to preserve CR.
People can then build on FOCIL and extend it to serve other purposes, but it’s confusing to tie FOCIL to scaling or even privacy considerations as it’s all a bit speculative for now.
The question should be:
- Do we think Ethereum’s current CR is good enough? I think there is broad consensus, and the answer is clearly not.
- And then, do we have many proposal that address this specific issue? I don’t think so. I think we have one, FOCIL, that’s been researched for years and now implemented by 4 clients. I would really like to know if there are specific concerns about the proposal itself though.
I also want to strongly make the point that there are no viable alternatives to FOCIL out there today, people can bring up ideas like MCP (one post, nothing on the actual way to merge different blocks together) or BRAID (promises of a revolutionary paper by Devcon Bangkok but really just some abstract slides from Max Resnick). In reality it’s unlikely MCP will “solve MEV”, and if they want to do so they introduce latency by using slow cryptography, or additional rounds of consensus, and in any case can’t prevent the last look problem and colocation with centralized exchanges.
So the only alternative there is right now is just letting 2 block builders completely control block contents and censor transactions without any safeguards.