…oak + e2e After #393 disabled automated CID delivery, the 4-hop round-trip scenario's HOP 4 — where bob bundles 3 source tokens with multi-hop histories — now throws INLINE_CAR_TOO_LARGE at the dispatcher pre-flight instead of trying to publish via IPFS. That throw is EXPECTED post-#393 and is orthogonal to the #391 fix being verified. Both the soak (`manual-test-roundtrip-391.sh`) and the e2e (`tests/e2e/issue-391-roundtrip.test.ts`) are updated to: - Capture the HOP 4 outcome instead of letting it abort. - Hard-fail if the throw is DUPLICATE_BUNDLE_MEMBERSHIP (#391 regression). - Soft-pass if the throw is INLINE_CAR_TOO_LARGE / "automated CID delivery is currently disabled" (#393 documented limit). Log the INFO, skip the receive-poll and balance reconciliation, and exit success — the #391 invariant is verified by the absence of the duplicate-bundle error. - Run the balance reconciliation only when HOP 4 actually delivered (the SUCCESS path). E2E also switched to `transferMode: 'conservative'` to match the working sibling `uxf-send-receive.test.ts` pattern; the instant-mode USDU path has pre-existing UXF orchestrator infrastructure issues (CBOR uint64 overflow on SMT path bignums, sibling header lines 35-58) that are independent of the #391 fix being verified here. CLI soak `manual-test-roundtrip-391.sh` remains the primary verification of the user's actual instant-mode bug repro; the e2e is a deterministic in-process companion focused on the guard invariant.

Migrate SDK v0.7.2 to state transition v2 #393

Description

Metadata

Metadata

Assignees

Labels

Type

Fields

Projects

Milestone

Relationships

Development

Issue actions