47 sats \ 15 replies \ @elvismercury 10 Dec 2023 \ parent \ on: Inscriptions: the cure is worse than the disease bitcoin
With every block full, that 1tb hard drive you bought years ago and connected to your Raspberry Pi will still take years to fill up. Hardware costs are not the issue.
The mempool filling up and UTXOs turning to dust is the issue.
You made a good argument. You don't need the "hardware costs hurting decentralization" fiction to justify it.
fees were always going to go up, we are just early
reply
Agreed. I think it would have been better, all things considered, to have had some more time, so that 'real' btc use could have developed more, but it can't be helped.
reply
I am glad that ordinals / inscriptions happened for this reason. It's good that people get to see Lightning operate in a high-fee environment and learn about its deficiencies in real-time.
reply
Can you explain its deficiencies in a high fee environment like im 5?
reply
Opening and closing channels require on-chain transaction fees.
So suppose I want to on-board someone onto Lightning, with a non-custodial wallet such as Phoenix. I could send you $10 worth of sats to get you started, but a good chunk of that will be spent on the channel opening fee.
There are custodial wallets out there where you don't have to worry about channel management at all. These are far cheaper to use as a result, but of course you have to trust the provider. They are perfectly fine for smaller amounts. However, we might be seeing less of these as they are a regulatory target (see: Wallet of Satoshi, BlueWallet).
reply
I see how that could be a problem but isn't Stacker.news custodial also? not like we have to keep opening and closing channels here. Also since bitcoin is permissionless shutting down one lightning custodian probably means 2 more take their place. The real threat imo would be the custodian running off with your money but if its a small amount its not as big a deal as your life savings. ALSO can't I send bitcoin to a custodial wallet like cashapp and send it from there to Phoenix Wallet?
reply
The principal issue under discussion isn't custodianship and its security vulnerabilities. The issue is that to get funds onto an L2, channels need to be opened and eventually closed. If it's five bucks to make a channel tx, that renders migration to an L2 infeasible for a vast number of UTXOs.
reply
yeah but again let's say I have bitcoin on cash app which has lightning functionality. can't I send it from there to Phoenix for barely any money? So I pay wayy less fees?
SN is custodial. So we're all using SN's channels here and the cost of managing those channels is abstracted away from us.
A non-custodial wallet basically means you have your own channels, that nobody else can control. So if you spin up a Phoenix wallet and deposit funds into it for the first time, a channel needs to be created for you.
So kinda like this:
Receiver = Custodial -> No on-chain fees incurred
Receiver = Non-Custodial with no channels / inbound liquidity -> On-chain fees incurred
Once the channel has been created, you can spend from it at little cost. If you want to create a new channel, change a channel's capacity, or close a channel, that all involves touching the base chain.
Savvy Lightning users will do their channel management in low-fee environments to minimize costs. Open a channel when fees are low, and spend from it when fees are high. But there will come a time where fees are always high, and most people will be priced out of opening their own channels (unless we see some crazy new developments in the space that I'm unaware of).
reply
Not everyone in the world can afford a 1TB hard drive. Yet another way ordinals are pricing out the poorer users of Bitcoin.
deleted by author
reply
You can bet on it because everything is bounded, barring some additional fork: if blocks are as full as they can be from now till eternity, the growth of the chain will be far less than the growth of relevant technology that supports chain growth (hard drive tech, in this case, but also processor power and bandwidth.)
I mean, I am seriously good for several more years on the piece of shit Raspberry Pi that I bought several years ago for negligible money. You can run a high-performance node in the form factor of a deck of playing cards. And that's without using a single scaling optimization, that's a full node with the entire chain history. The present resilience is testament to the vision of the Core devs for these last years that decentralization has been prioritized so hard.
This sounds at this point like I'm nitpicking, but I'm just saying, there are legit philosophical and design tradeoffs to be made here, so we don't need to water down the argument with non-issues.
reply