spacebear

remaining TODOs per @nothingmuch : 1. (DONE) making the receiver's request body handling stricter 2. the same change to the sender 3. enforcing limits on payjoin-cli receiver, which currently uses...

We just need to bring in the BDK test from payjoin-ffi into rust-payjoin. After https://github.com/payjoin/rust-payjoin/issues/790 we will no longer really need to test the `payjoin-ffi` wrappers directly, instead we can...

Closing this as it has been mainly addressed by #712 #753, with some small follow-ups still needed.

Adding to 1.0 milestone not necessarily because I think we should use strong types, but we should make a decision for 1.0 either way.

> It may be possible to maintain the public FFI surface for simple primitives (PSBT strings, Satoshi u64, plain URLs) in the 1.0 cycle and compensate with stricter runtime validation....

We must stay on 1.63 to respect the MSRV in the wider rust bitcoin ecosystem (e.g. [rust-bitcoin](https://github.com/rust-bitcoin/rust-bitcoin?tab=readme-ov-file#minimum-supported-rust-version-msrv), [bdk](https://github.com/bitcoindevkit/bdk?tab=readme-ov-file#minimum-supported-rust-version-msrv), [rust-lightning](https://github.com/lightningdevkit/rust-lightning/blob/0fe51c5af0fe9399fce6a45b8cc6ba144e45a568/CONTRIBUTING.md#contribution-workflow)).

Converting to Draft because I'm not sure whether PJ_STORAGE_DIR needs to be set explicitly? cc @nothingmuch

This change also requires an up-to-date docker image for `payjoin-directory` so marking this as blocked until then.

Closing this per #1190

Closing this and #596 as no clear direction was decided on.