Nick Szabo  🔑

Try to use that 9 sats to send a micropayment to someone on layer 1.

I don't need to. On-chain is not a good UX. Fees are calculated on UTXO data, it's handicapped from the get-go vs. payment channels. If you can make payments and not have to deal with UTXOs & data consolidation, that's superior for privacy and fungibility.

LN doesn’t allow you to move funds that are less than the cost of closing a channel. So you can’t even if you wanted to.

When you fund a channel, you put satoshis into a reserve to pay the channel closing. It's cheap to close it because it's SegWit & only one UTXO. You can batch LN channel closings together as well in the future; have multiple inputs/outputs during opening & closing transaction

Each channel contains a reserve which is unspendable and typically takes about 2% of the channel’s capacity. You must spend an amount matching that reserve to make receiving possible. Channel security is reserve + timelocks. http://lightning-wallet.com/using-lightning-wallet#using-lightning-wallet …

I don’t have anything bad against LN, it’s a very interesting software eng challenge & it will definitely be used in specific application with medium size channels. However it’s bad solution for BTC worldwide micropayment scaling solution. I was trying to point that

You just contradicted yourself. You said LN will be used for medium size channels, but that it's bad for micropayments. This makes no sense, because LN excels at micropayments. It is unicast, not broadcast. You are only broadcasting payments between you and the receiver.

I didn’t, I said LN can be utilized with network that have small to medium NUMBER of channels. But If you use it as a global network with millions/billions of channels the routing complexity is exponentially high & current unsolved