Availability is not a problem, until it is. Right now there are three public coordinators that could be compromised, attacked or shut down. Even when I am sure new coordinators would pop up in case of mayor disruptions, and I am also sure liquidity will eventually find its way to the next coordinator, the process is slow and painful because requires users to choose a new coordinator.

Running a coordinator in each Wasabi client would make the whole system more resistant and would remove the easy targets by distributing the load of coordinating transactions.

Reply to this note

Please Login to reply.

Discussion

No doubt making clients potential coordinators is a step toward better decentralization. But that’s not my question, which is: is lack of coordinator decentralization preventing people from using wasabi who otherwise would?

I think so, yes. Because the lack of coordinator decentralization ruins the UX by forcing users to research how the whole thing works, find a coordinator and configure the client to use it. Having decentralized coordinators would allow us to select the coordinator for the next round automatically making the feature immediately available after Wasabi installation for everyone.

Why don't you select randomly already? If it's because you don't feel it's appropriate to give implicit trust to the coordinators, I understand that, but then surely that wouldn't change if it was decentralized?

That was the very first idea that came to my mind, but then the question is once again: how to prevent someone from generating a million identities?