docs: update content on paymentpointers.org

[melissahenderson]

Revise the paymentpointers.org documentation so that it accurately reflects the relation between payment pointers, wallet addresses, ILP/SPSP, and Open Payments.

Background:

Payment pointers were originally introduced as an Interledger RFC: https://github.com/interledger/rfcs/tree/main/0026-payment-pointers

The RFC became its own website: http://paymentpointers.org/. The sidenav in the Interledger Developers site links to paymentpointers.org.

Other than a tool to convert a payment pointer to a URL, the content remained largely the same as what was in the original RFC until 2020. Then, Adrian made a number of updates to work in Open Payments. Max pointed out that the site doesn’t mention anything about Interledger/the protocol and all references lead to Open Payments.

A URL can be both a payment pointer and a wallet address. Payment pointers are Interledger-only. Currently, wallet addresses (Open Payments) are also payment pointers because Open Payments only works with ILP. If/when Open Payments works with more than ILP, you could have a wallet address that doesn’t support Interledger (and as such, is not a payment pointer).

There’s also the issue of terminology. For example, Chimoney and GateHub call the URL a payment pointer. Fynbos calls it a wallet address.

Payment pointers are technically built on SPSP and wallet addresses on Open Payments, but the two are competing application layer protocols in the ILP stack. If you query a URL that’s both, you call with SPSP to get a payment pointer response and with application+json to get a wallet address response.

Other info: The site states, “A Payment Pointer resolves to a URL (with the https scheme) that can be used to discover the Open Payments endpoints for interacting with the account. Using the Open Payments protocol the counter-party can initiate a payment to or from the owner of the Payment Pointer.” This could be construed to mean that payment pointers are definitively tied to Open Payments, except someone could choose to use payment pointers in some other way (e.g., Interledger/SPSP).

[melissahenderson]

PR #57