AMP: Atomic Multipath Payments. The sender creates MULTIPLE routes through different payment channels to the receiver. This is atomic, in the sense that either ALL routes succeed, or NONE.
Keysend: The sender does not need to ask the receiver for an invoice [hash of a pre-image], but the sender encrypts a secret to the receivers public key, and uses this cyphertext to build the HTLC route. The receiver can then decrypt the cyphertext to reveal the preimage.
Regular lightning requires the receiver to generate a secret, hash it, put it in the lightning invoice, and give that to the sender to construct the route.
Keysend is as described above that the sender creates the secret and encrypts it to the receivers public key. Thus the "invoice" is a static public key.
LNURL encodes a static domain that links to the node, the sender then talks to the receivers server to get the invoice.
"Keysend is being deprecated in favor of AMP. Keysend was meant to be just a hold over until AMP was rolled out..."
@maxhillebrand @adam Roy at Breez had explained to me a few weeks ago that what he called "static invoices" would be a way to replace it as well with an open ended invoice generated by the receiver, thus placing the preimage back in the hands of the receiver instead of with the sender as it is with keysend.
That was my understanding from Roy's comments, but maybe I misunderstood him.
If the receiver generates the pre-image, then the sender only gets it if the payment went through successfully.
This is somewhat of a "proof of payment".
Right now, the same pre-image is used for every hop in the route, so even routing nodes can provide this "proof" [thus it's not really a proof after all...]
However, with Taproot / Schnorr / Adaptor Signatures, each node has a different "pre-image", thus only the sender will get the "ultimate pre-image", so then it's a proof.
Cannot find anything other than the above described flow in the optech archive...
Intended for all stake holders of podcasting who are interested in improving the eco system