Oddbean new post about login | logout regular reminder that lightning *is* a “shared utxo” protocol that uses covenants (and multisig) to scale bitcoin payment tx thruput most of the commentary against it points out that it does not scale number of holders in relation to # of utxos. in my mind this is a good sign of the protocol’s success (it shipped and people use it and they found drawbacks with the MVP). It’s also a sign of bad messaging around the applicability of 2-party accounts to certain payment situations (venmo for example) sometimes it is useful to have multiparty channels; i’d argue that liquid is one such example of this. federated mints are another one; note that both of these multiparty systems require consensus algorithms to function. fedimints were using BFS, (iiuc they just changed to a new one). liquid uses “nakamoto consensus” aka a blockchain to keep track and agree on shared balance states Lightning was shipped as the minimum viable product to scale tx thru-put and it succeeded at that. it did this with minimalist, 2-party design as consensus from 2-parties is the simplest case. this meant it could avoid tricky consensus questions and focus on privacy + throughput instead shipping multi-party utxos at scale will require a solutions for managing consensus amongst stakeholders in the “shared utxo”; see Ark and already existing federations for examples of contending with this problem 🫡
ahem which covenants do you speak lady? what i speak against on it is that it is source routed and that's baked into the HTLC payment transfer design and that's the reason it has weak liveness no, lightning does not need a consensus algorithm, it is pure two party, peer to peer on which planet do you live milady? have you ever studied distributed systems theory? i guess not, you worked for Blockstream
i don't know how to say in how many ways you are completely incorrect in what you are saying about lightning but you can read my other comment and i suggest you go read up on distributed systems theory please for the love of god
"i don't know how"... You could have stopped right there. Both of your comments make no points but only insult the OP and Blockstream. But you don't offer any solutions or any reasons why she is wrong, etc.
is this the kind of avatar that blockstream is paying to push on the world or is she a ronin now or something? nostr:nevent1qqsfspetwhn4yz5kmu73r949203w499x023dxkt5zultssse2g0ezlsppemhxue69uhkummn9ekx7mp0qyt8wumn8ghj7etyv4hzumn0wd68ytnvv9hxgtcpr9mhxue69uhhyetvv9ujuumwdae8gtnnda3kjctv9uq32amnwvaz7tmjv4kxz7fwv3sk6atn9e5k7tc5jrn46
one thing i think bitcoiners lose sight of is that we have successfully launched the largest and most decentralized layer-2 payment network (after credit cards) there is no other ecosystem that has achieved this. most L2s on other ecosystems dont even get close to the *Liquid* federation’s level of decentralization. they just use centralized servers and a manually coordinated multisig that brings with it a real cost: decentralization is expensive. duplication is basically the bedrock of decentralization but that also kills “efficiency of liquidity”, which is basically what every post-LN improvement is trying to fix upgrades happen in phases: we try going left, discover problems and course correct to go more right etc I’d expect the next iteration of “shared utxo” protocols to be more centralized in terms of liquidity, with the tradeoff of far more consensus design work (who all needs to agree on the current state of the world for us to move forward?)
Re. next shared utxo: maybe a stock settlement system?
whats a “stock settlement system”? Guessing here but closest thing this sounds like that currently exists is Liquid’s asset issuance platform (securities shipped and held as stuff in your Liquid wallet)
I don't know what fedimint is but that's an interesting note
https://fedimint.org Chaumian ecash pegged to bitcoin. But with a modular chassis so you can do other interesting things too.
Thanks G
Well said, I think lightning will be the decentralized glue for all the less decentralized L2s
This makes sense to me. The implication seems to be that only a few large stackers will be able to be sovereign. Everyone else will be hodling in some form of custodial system. I’m optimistic that the community will eventually figure out a better and more decentralized layer.
Yeah, we're just going to need soft forks most likely
is lightning decentralized though?
yes, of course it is. we have a strict decentralization verification process. we guarantee decentralization by having every LN transaction’s route be chosen by a single server, kept in a closet at your mom’s house o.O
A paper from 2020 co-authored by Christian Decker shared that 10% nodes hold 80% liquidity: LN: A second path towards centralization of the bitcoin economy - https://arxiv.org/pdf/2002.02819.pdf Liquidity in 2024: https://mempool.space/lightning/nodes/rankings/liquidity Maybe someone could write another paper with present stats including use of custodial wallets
The higher the layer the more centralized things will be as systems are rebuilt. But the bedrock will be the decentralized base layer that is the bitcoin protocol. Scams will happen as people discover ways to rehypothecate but I say again the base layer will remain stalwart
Thank you for this, I am trying to learn this vast part, it is ultra important to Bitcoin. I am about 200 hours behind in learning.
It seems to me narratives around lightning are changing a little right now. When it was newly reckless it was the closet cypherpunks Bitcoin scaling system meant to be run on the merest hardware for individual sovereignty. As we learned that some of its trade-offs made it impractical for even fairly advanced tinkers we watched as professional lightning nodes emerged. Nodes definitely NOT running on raspberry pis and locking up 10s-100s of btc into htlcs and these nodes were the custodial wallet providers, other LSPs and the individuals willing to work a part time job managing their nodes. So what is the narrative shift? I think #Aqua is a great example. We are seeing professionals combined lightning with liquid and probably eventually other later twos and LSPs and other services to provide and at least semi sovereign product for users who couldn't even begin to run a lightning node. It is beginning to look like lightning is going to be more what liquid was sold to be in the beginning. A way for professionals to move value back and forth between large endpoints. While liquid may end up stealing a little of the stage from lightning when it comes to self sovereign simple value transfer. And the combination is really strong. I think once we start seeing some of the other layer 2 frameworks coming up we can go nowhere but up from here.
Fedimint switched recently from HBBFT to Aleph BFT.
While I understand what you mean by "channel" I believe calling Liquid a channel would greatly confuse people, especially non-experts.