PDK Dev Call Update - July 21

[arminsabouri]

PDK Dev Call Updates – July 21st

Armin

• Finished the Liana database refactor and implemented database tests.
• Completed the separation of fee application from signing Seperating fee application from signing #870 and integrated the changes into Liana.
• Maintained two threads of communication with the Liana team: one regarding support for foreign inputs, the other focused on UI/UX feedback.
• This week: refactoring how payjoin status is derived, removing the redundant status endpoint, and cleaning up error handling logic in Liana.

Ben

• Finished setting up a reliable Nix environment for Liana with external help.
• Worked on the FnMut closure refactor in rust-payjoin Use FnMut closures for receiver typestate checks #883
• This week: shifting focus to UI improvements and user-facing error handling.
• Liana’s GUI framework has a huge learning curve.
• Coordinating with Liana developers and Bitcoin Design contributors to walk through the payjoin flow and collect feedback.

Chavic

• Focused on UniFFI test FFI fixtures. Only a few remain.
• Plans to write a README to help others begin contributing to the project.
• After importing the rest of the fixtures, will move on to repairing broken tests.

Dan

• Investigated compatibility concerns between Bull Bitcoin and Cake Wallet. Goal is to confirm tools work in real-world usage.
• How do we fix this? First step: ensure tools work consistently in local development setups.
• Removing the UniFFI feature flag and the ffi parent mod in payjoin-ffi would reduce maintenance overhead.
• Broke down maintainer responsibilities and began recategorizing them. This will be documented (somewhere?)
• Refactored sender API. The current hierarchical structure needs to be decomposed Abstract common sender functionality, not heirarchy #884
• Supported BDK CLI integration. Efe had issues using extract_req to identify payjoin version.
• Noted problems with the sender builder and session handling flow What's up with send::v2::Sender #809
• Met with Yash in Boston to align on collaboration.
• Reviewed the PartialEq PR Add PartialEq/Eq to errors for easier comparison #768 and continuing to reivew other PRs
• Continued legal coordination for drafting organizational documents, with input from founders of other successful SDK projects.

Nothingmuch

• Rebased the Tokio + Rust-TLS ACME refactor. Reduced it to 3 commits. Aiming to open the PR soon.
• Follow ups: Integrating clap and the config trait; reassessing ergonomics of running payjoin-dir. These follow ups will be documented issues and discussions.
• Reviewed the FnMut PR. Wants to collect feedback from others Use FnMut closures for receiver typestate checks #883 .
• Will Schedule a follow-up NixOS call to assist with Dart-UniFFI integration.

Spacebear

• Finalized and merged the UniFFI-Dart PR Uniffi dart #829
• Tested payjoin state across Cake Wallet and Bull Bitcoin. Found that payjoins failed bidirectionally. Bull Bitcoin devs had successful local tests, so the issue may lie with Cake. Without us running cake locally on our machines and testing its hard to tell.
• Investigating how to standardize testing across implementations.
• Sessions work on the happy path, but error handling is lacking when things fail.
• Worked on running Bull Bitcoin Mobile locally with the UniFFI-Dart version instead of the rust dart bridge.
• Started writing the talk for Bitcoin++.

0xZaddy

• Rename _danger-local-https to _manual-tls Rename _danger-local-https to _manual-tls #763
• Scrubbing logs: issue Add documentation builder to pre-commit hook recommendation #820.

Shehu

• Continued migrating sled to sqlite3 in payjoin-cli Migrate payjoin-cli Sled db to sqlite3 #867
• Addressed initial review comments and is now requesting additional feedback.

Zele

• Opened PR Remove unnecessary payjoin-ffi dependencies #866 to replace Linux-specific flock with a cross-platform alternative for Unix-based systems.
• Followed up by adding macOS and Windows CI tests Add macOS to CI test matrix #888
• Encountered blockers: lint checks fail in pre-commit hooks due to warnings about large enum variants.
• This week: performing code reviews and picking up open issues.

Blu

• Opened a followup PR to #758 to add the "Feature Request", "Good First Issue", and "General" feature templates.
• Address feedback for python FFI Nix devShell, after reviewing feedback I realize I need to spend some time diving deeper into Nix flake best practices to ensure my modifications are stable and useful long-term.
• Start writing companion article to spacebear's btc++ presentation.