Do you disagree with his numbers and prospects for UTXO ownership?
Discussion
Not sure which numbers you’re referring to, but I think I agree with the *need*, just strongly disagree with the idea that we have any great solutions. From where I sit, the ideas for scaling using covenants often don’t provide all that much scale, and almost always make other tradeoffs.
IMHO one of the more compelling ideas is timeout trees (which sadly was proposed after people had kinda given up on CTV). But it makes a huge tradeoff - if you get screwed it costs you 10-20x more to FC than Lightning. Worse, an operator can create a “ghetto” of poor users and screw them all to steal from them. *but* it gets you great optimistic-case scaling!
Most things end up looking like that, and it leaves me pretty unexcited…except that the research is moving at a clip! I’m excited to see what the research comes up with, in the mean time I’m still building lightning.
I share your skepticism about how much covenants can scale on-chain throughput. James makes a good point about CTV alleviating the thundering herd problem. But if fees are permanently higher than what people can afford it won't help.
No one will ever use congestion control as defined by CTV.
I agree with you that all proposed L2 solutions based on covenants are horrible (except for Mercury statechains but it's not fully trustless so I guess that makes everybody ignore them?).
I don't understand the last paragraph. What is this research you're talking about? I want to be excited about something too.
There’s a bunch of research on the limits of things - Shielded CSV/rollups (though want to see the limits of a CSV thing if we build a version you can do lightning on top of), Ark/timeout trees is a cool set of ideas that feels like it could be extended, etc. The results certainly don’t point to us having any silver bullets, but I think some of those things probably with some more tweaks may give us a nice direction. But certainly want more analysis.
By the way, do we really any prospects of creating a trustless decentralized L2 on Bitcoin ever? Isn't it true that Ethereum has had all the ability to make any covenants they want for years and after so much time trying to find a scalable solution the best they got was zk rollups that centralize everything in the hands of a company and require a ton of space anyway that has to be acknowledged by the chain consensus?
In other words, even if we had all the opcodes we wanted by magic we would still be just as bad at scaling as Ethereum is currently?
Yea, I mean the ETH zkRollups are super centralized but to some extent that’s a consequence of these things being hard to build so they need an “escape valve” in practice. There’s not a lot of motivation to remove that but I assume it’ll get removed eventually…. just may be a while.
I do think it’s somewhat informative that ETH only built *rollups*, which certainly don’t have a huge scalability multiple themselves (doubly so for payments, though it’s important to note ETH is trying to scale computation, not payments).