pull down to refresh

Context: #1559774

Why you don't just ask for a lightning address instead? or you really want to know which node the invoice is coming from?

It seems like @AGORA thinks I’m asking for lightning invoices instead of lightning addresses because I “want to know which node the invoice is coming from.“

What do you think? Would a lightning address be better?

Yes10.0%
No90.0%
20 votes \ poll ended

In general, no addressing solution beats an addressless solution. But BOLT-11 isn't addressless.

You could do blinded path on a BOLT-11 with bLIP39 too, LND supports it, but I'm not sure what other implementations do. Some caveats:

  1. It's directly noncompliant with the BOLT-11 spec because no s tag
  2. You need to be able to decode b tags and these are variable length
  3. It makes your invoice YUGE.
  4. In the field, especially in my shaky hands that break things, those seem to often fail.

I think @anon + bolt11 is fine, as long as you also @delete?

reply
127 sats \ 14 replies \ @ek OP 1 Sep

Oh, yeah, I could recommend @delete, too! I'd only considered PGP encryption support in an update. From @ccbank's bio:

Future versions might use PGP for encryption to hide the recipient's node pubkey in the lightning invoice. Only the payment hash and amount need to be public for payment verification.

@delete would def mess with payment verification, but I'm not sure how important that really is. It's still trust-me-bro on every trade.

reply

If either Goog finally pushes xwing into SubtleCrypto as they said they would months ago, or Apple stops dicking around with restricting custom calls on Webauthn, there will be a path to having private messaging in the browser despite people yoloing plugins.

Until then, there's only "private" messaging, aka leaky shit, or a massive capture-now-decrypt-later gap.

Edit to your edit: why do you need PUBLIC payment verification?

reply
127 sats \ 11 replies \ @ek OP 2 Sep
why do you need PUBLIC payment verification?

The idea was reputation but not sure it’s necessary

reply

I don't think there is a transparent way to truly account the fulfillment side due to LN sender privacy anyway, so it's always going to have some trustmebro.

What's more interesting to me is: whatcha gonna do with all them CCs stuck in a bot account? Even if you save up to 3M, transfer territory, pay the forever bill, transfer back, that's a one-time thing.

reply
127 sats \ 9 replies \ @ek OP 2 Sep
watcha gonna do with all them CCs stuck in a bot account?

Treasury is capped at 100k credits for now. I just want to buy CCs cheaply for territory rent. I don’t want to pay the forever bill.

reply

right so your demand is basically 50k/mo

reply
127 sats \ 7 replies \ @ek OP 2 Sep

Plus any demand from other founders who want to pay their bill cheaper, too.

The amount of trust involved (transferring territory to @ccbank and trusting it to transfer it back after payment) sucks. Maybe I’ll find some way to reduce the trust required with DLCs and an independent oracle (you’ll get paid on-chain if the bank doesn’t give you the territory back), but idk. That’s future stuff haha.

It’s an interesting problem though.

Usage: @ccbank sell <invoice>
  Exchange credits for sats. <invoice> must be a mainnet lightning invoice
  with an amount that is valid for at least 24h.
reply
264 sats \ 3 replies \ @Murch 1 Sep

Doesn't a lightning address just encode where to find the node to ask for an invoice from?

reply
235 sats \ 2 replies \ @ek OP 1 Sep

Yes, but worse: The HTTP server serving invoice requests is usually run by a trusted third party, so it’s additionally bad for privacy and security compared to a lightning invoice.

(Of course lightning addresses are still useful for other reasons.)

reply

Clink fixes this

reply
278 sats \ 1 reply \ @nullcount 1 Sep
I want to know which node the invoice is coming from.

A BOLT11 does not have to necessarily reveal this with certainty. Phantom node payments allow lightning nodes to generate invoices that can be paid to one of multiple nodes.

This is a feature in LDK that is marketed towards enterprise load-balancers, but it also can be used to get some deniability that you received a transaction. Other implementations could add this phantom invoice feature trivially (but they won't), or you can DIY it today with one LLM prompt.

Would a lightning address be better?

Lightning address is just a protocol to request BOLT11 invoices (which could be phantoms BOLT11 invoices)

Using lightning address requires DNS, but it doesn't have to be your IP-address hosting it. Likewise with BOLT11 phantoms, you can loadbalance/distribute the lightning address server itself.

Nothing stops you from hosting the lightning address server over a .onion (well, except that none of the wallets will actually query it correctly) Be the change you want to see!

reply
12 sats \ 0 replies \ @ek OP 1 Sep

Oh, cool, didn’t know about phantom node payments!

reply
reply
2 sats \ 1 reply \ @ek OP 2 Sep

your federal agent loves that you think about him all the time and police yourself

reply
all the time?

lol, no - i barely have time to think about myself and my immediate surroundings; i do care about the watchers' well-being, many of them are heavily intoxicated and constantly scratch themselves ... (parasites);

reply

bolt12 with blinded paths is better. coinos.pro has them

reply
2 sats \ 3 replies \ @satoshiplanet 1 Sep -30 sats

No. Lightning address is convenience, not privacy. @nullcount already explained phantom invoices, that's actual privacy. @ek nailed the HTTP/third-party issue. Address adds exposure, not protection, use the right tool for the job