Oddbean

In other words: you're going to struggle to get that thing to have some network-effect and then suddenly the company will disappear and everything will vanish. It would be much better if the scaling solution was done collectively by all people who want to promote Bitcoin adoption with no immediate financial concerns in mind, and everybody agreed on what to work on and worked together towards that, without being led by a company. Companies are welcome to join, but they shouldn't lead.

The Lightning Network was beautiful for a while because it met all these requisites, and it had a chance of working, but unfortunately that wasn't meant to be. Drivechains could also meet these standards, as they would necessarily have to be tacitly approved by the community at large in order to work, but I don't know what else. Maybe Liquid and Rootstock get enough lindyness on them after so many years that they become dissociated from the companies and start to be perceived more as public goods, and that causes them to be more used? I don't know. Mercury statechains could also be a very nice neutral scaling solution if more than one company started to run a signer server.

I'm trying to get my head around the various designs.. And it seems to me that anything with a clear icon has an unclear future. What do you think about Lightning Channel Factories… Isn't there hope? https://image.nostr.build/ef84276e62818b034da096a444c428dcd3b1401bbe8e5cdca1188344d0a47089.jpg

I think it doesn't help much: Lightning is broken at a fundamental level with HTLC settlements (see @htlc_nbot), fake HTLCs and the requirement of immediate onchain settlement, and the way it manifests itself is through the absurd number of force-closes that happen daily (see @LN_FC_Bot). Channel factories, timeout trees and other ideas just allow you to create more channels, but they will keep closing. Besides that it might add so much complexity it will probably 5 years to implement and then yield another 5 years of bugs that cause channels to close more than required. For example, bolt12 it relatively simple, makes no change to the nature of channels and is taking almost 5 years already just to implemented with no real-world usage. See also: nostr:naddr1qqyrqdr989jnsvf5qyghwumn8ghj7enfv96x5ctx9e3k7mgzyqalp33lewf5vdq847t6te0wvnags0gs0mu72kz8938tn24wlfze6qcyqqq823cjxzu6c

Thank you for phrasing the question as "least bad". Drivechain and Mercury Layer (statechains)(hopefully with some APO magic that allows statecoins to live forever). But we still need something for micropayments that can be more trusted. Maybe if these custodial ecash mints could settle among them without using Lightning. Or maybe just using Lightning in a low-fee drivechain will work even though it's ugly and dirty.