-
Notifications
You must be signed in to change notification settings - Fork 0
Open
Description
Description
Recent discussion highlighted ambiguity around using client=ark versus wallet=<walletname> in payment requests/metadata, and what guarantees each implies.
Key points:
wallet=<walletname>(e.g.,wallet=noahwallet) is effectively proprietary and can assume Ark-only handling (i.e., can guarantee no Lightning invoice is needed).client=arkis intended to be more interoperable, enabling wallets to detect whether a destination ASP is supported.- If
client=arkis used and the destination ASP is not supported, the sender may need to fall back to providing/using a Lightning invoice. - This could be important for scenarios like offline payments and cross-wallet payments, but it requires a clear spec.
Context
- There was a request to consider making breaking changes early if moving from
wallet=noahwallettoclient=arkis desired, to avoid doing it later. - It was clarified that
wallet=noahwalletcannot be replaced outright withclient=arkbecauseclient=arkcannot guarantee Ark-only settlement and may require LN fallback. - Agreement that a specification is needed before implementing or relying on this behavior.
Expected Behavior
- Define a spec that clearly states:
- The intended semantics of
client=arkvswallet=<name>. - What a sender must include when using
client=ark(e.g., ability to provide LN invoice / LN fallback path). - How wallets should determine “supported ASP” and how that is represented/verified.
- Whether and how this enables offline payment flows.
- Backward compatibility expectations and whether any breaking changes are planned.
- The intended semantics of
Reactions are currently unavailable
Metadata
Metadata
Assignees
Labels
No labels