Problem
SendspinKit still needs library-side seams to let conformance use the real SendspinClient instead of a bespoke protocol client.
The conformance rule should be strict here: if the library cannot exercise a scenario, the test should expose that gap. The adapter should not fill in missing behavior by manually constructing/parsing protocol messages.
Why this matters
For server-initiated and binary-verification scenarios, conformance needs a minimal way to:
- hand an already-accepted connection into the real client implementation
- observe the library's actual handshake / stream / disconnect behavior
- verify artwork/audio delivery without reimplementing the protocol in the adapter
Local exploration also showed that once conformance is routed through the real client, actual library issues become visible instead of being hidden by bespoke adapter code. That is exactly what we want.
Scope
Keep this limited to the minimum surface needed to make testing possible. Do not use this as a reason to add unrelated new features.
Tracking
Problem
SendspinKitstill needs library-side seams to let conformance use the realSendspinClientinstead of a bespoke protocol client.The conformance rule should be strict here: if the library cannot exercise a scenario, the test should expose that gap. The adapter should not fill in missing behavior by manually constructing/parsing protocol messages.
Why this matters
For server-initiated and binary-verification scenarios, conformance needs a minimal way to:
Local exploration also showed that once conformance is routed through the real client, actual library issues become visible instead of being hidden by bespoke adapter code. That is exactly what we want.
Scope
Keep this limited to the minimum surface needed to make testing possible. Do not use this as a reason to add unrelated new features.
Tracking
SendspinKitSendspinClientaiosendspin -> SendspinKitconformance slice and record the remaining failures/passes