Skip to content

Transcript drop: tombstone delivery gaps (post-gc snapshot, push-only remote re-evaluation) #230

Closed

Description

@tonibergholm-codento

Follow-ups from PR #226 (sc transcript drop, #212). Both are documented limits in the ADR-0038 addendum, not regressions.

  1. Tombstone not sent after the dropper's gc removed its snapshot. The sender selects tombstones through the snapshots of the transfer set. gc keeps the tombstone but a collected snapshot is never in that set, so a stale peer that still lists the body is not told about the drop. Options: send every retained signed tombstone on every transfer (one small object per drop per fetch), or match against the raw advertised haves.
  2. Previously rejected tombstone is not re-evaluated on a push-only remote. A remote that received a signed tombstone by push before trusting the signer holds the object, so push never re-sends it and index_incoming never re-runs. The drop stays inert until the operator re-sends it. Options: re-evaluate held, unrecorded tombstones when [signers] trusted changes (or on gc/reindex), or force tombstone retransmission on push.

Acceptance: a stale or late-trusting peer converges on the drop without operator workarounds, with tests for both cases.

Activity

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