Skip to content

Follow-ups from the #227 (max object size) review: wiring test, stale comment, decode_owned edge tests #232

Closed

Description

@tonibergholm-codento

Non-blocking items from the renewed Opus review of #227.

  1. Wiring is untested: object_boundary_exactly_cap_accepted_cap_plus_one_rejected calls check_object_len directly, so deleting the check from the serve_session PutObject arm (wire.rs ~1214-1221) or from WireClient::get_object (stdio_transport.rs ~125-129) would pass every test. Add a scl-repo integration-test binary that sets SC_MAX_OBJECT_SIZE=2MiB before the first read (pattern: crates/core/tests/max_object_size.rs) and sends a 2 MiB + 1 PutObject through serve.
  2. Stale comment at wire.rs ~86-91 (chunk_from): 'the receiver refuses a frame above the object cap' is no longer true (frame limit is cap + 64). Chunks stay at cap - 1 so frames also fit receivers built before feat: operator-configurable max object size (--max-object-size / SC_MAX_OBJECT_SIZE) #227; say that.
  3. decode_owned test gaps (wire.rs ~1518): zero-length body, and a non-PutObject request passing through.

Activity

  1. added a commit that references this issue on Oct 20, 2026
    b7b1cc3
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    enhancementNew feature or request

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions