People aren't using custodial lightning wallets to save on channel opening fees
Not so sure about this, custodial does have economies of scale and sunk costs.
The tinkering user may pay more in custodian fees over time, and a volume user definitely will, but the upfront cost is practically nothing. Users generally have a very high time preference and will gladly pay 10k sats over a year than 1k sats up front.
they don't want to have to open and close channels
Primarily because they see the cost
they don't want to think about liquidity
Phoenix is a good example of idiot-proof automation which is why its done well, but the biggest complaint users have is about costs.
they don't want to run a lightning node
This is likely the biggest factor, and Lightning is a cost center if you're offline. Custodians are always online. Phoenix/Zeus are so expensive because you can't earn money while you sleep, and a great example of why mobile nodes are a dead-end because they'll always be a cost-center not a tool for revenue.
So refine my earlier point, ABSOLUTE COST is not a barrier to adoption... but up-front or mental costs may be.
Channel factories don't solve that any better than batch opens, nor do they absolve the trust and centralization of custodial "graduated" on-boarding.
I'm curious how different the trust assumption is between a single channel with Phoenix LSP and something like Spark.
They're the same because they face the same underlying physics of the chain, the techno-babble Fake L2's use is an effort to distract people from this and why they are scammers.
They have to scam the user so that they be consistent when scamming the regulator, they're big companies that don't want to attract attention as a custodian, and they don't provide any real services they can monetize.
Phoenix bootstraps your liquidity with a trusted LSP credit (Lightning.Pub does this too). During this time it is completely trusted. It has to be this way because if you can't afford to hit the chain you can't afford the security of said chain. They do not feel the need to scam the regulator because what they're custodying is a service credit, not a financial service.
Once you can redeem that service credit for a proper LN channel, it becomes trustless because of the unilateral exit ability since Lighting is chain anchored.
Spark is similar, it's trusted until you exit to a real channel.
Anything above the chain has an online-ness requirement, that doesn't make it not trustless, because you still control whether your defending your claim or not. It's an additional security trade-off you're making to coordinate above the chain, not trust.
Any L2 system will have this trade-off, and there's also no such thing as off-line payments on L2's despite the claims of many a scammer or script kid. You need to be online to communicate with the chain to redeem and defend any claims you may have on layers above it.
Not so sure about this, custodial does have economies of scale and sunk costs.
The tinkering user may pay more in custodian fees over time, and a volume user definitely will, but the upfront cost is practically nothing. Users generally have a very high time preference and will gladly pay 10k sats over a year than 1k sats up front.
Primarily because they see the cost
Phoenix is a good example of idiot-proof automation which is why its done well, but the biggest complaint users have is about costs.
This is likely the biggest factor, and Lightning is a cost center if you're offline. Custodians are always online. Phoenix/Zeus are so expensive because you can't earn money while you sleep, and a great example of why mobile nodes are a dead-end because they'll always be a cost-center not a tool for revenue.
So refine my earlier point, ABSOLUTE COST is not a barrier to adoption... but up-front or mental costs may be.
Channel factories don't solve that any better than batch opens, nor do they absolve the trust and centralization of custodial "graduated" on-boarding.
They're the same because they face the same underlying physics of the chain, the techno-babble Fake L2's use is an effort to distract people from this and why they are scammers.
They have to scam the user so that they be consistent when scamming the regulator, they're big companies that don't want to attract attention as a custodian, and they don't provide any real services they can monetize.
Phoenix bootstraps your liquidity with a trusted LSP credit (Lightning.Pub does this too). During this time it is completely trusted. It has to be this way because if you can't afford to hit the chain you can't afford the security of said chain. They do not feel the need to scam the regulator because what they're custodying is a service credit, not a financial service.
Once you can redeem that service credit for a proper LN channel, it becomes trustless because of the unilateral exit ability since Lighting is chain anchored.
Spark is similar, it's trusted until you exit to a real channel.
Anything above the chain has an online-ness requirement, that doesn't make it not trustless, because you still control whether your defending your claim or not. It's an additional security trade-off you're making to coordinate above the chain, not trust.
Any L2 system will have this trade-off, and there's also no such thing as off-line payments on L2's despite the claims of many a scammer or script kid. You need to be online to communicate with the chain to redeem and defend any claims you may have on layers above it.