Annotate rewrite tracker against current branch state

4b63c50→main

Sessions

aac6eaf4698eView transcript

Changes

1

• docs

  • Mrewrite-issue-list.md+62/-31

66 unmodified lines

1. Batched bootstrap can claim tag refs were pushed when no tag ref was created

Status: open
Status: done

Problem:

• In the batched bootstrap tag phase, FetchPack returning git.NoErrAlreadyUpToDate can skip tag creation entirely even when the tag ref is absent and only the tag object is already reachable.

2. Duplicate target mappings are silently accepted

Status: open
Status: done

Problem:

• Multiple mappings to the same target ref overwrite each other silently.

3. Mapping normalization accepts inconsistent ref kinds and partially-qualified refs

Status: open
Status: done

Problem:

• Invalid combinations survive normalization and fail later with misleading errors.

4. Sideband preference is backwards

Status: open
Status: done

Problem:

• The current code prefers sideband before sideband64k.

5. Pack reader leak in bootstrap batch loop

Status: open
Status: partial

Problem:

• Batch loop fetches a packReader without robust close discipline on all error paths.

6. Protocol v2 tag fetches request include-tag without capability gating

Status: open
Status: done

Problem:

• v2 request building sends include-tag without first checking server support.

7. OAuth refresh failures are swallowed and stale tokens are reused

Status: open
Status: done

Problem:

• Token refresh failure degrades into later auth failure with poor diagnostics.

8. statsCollector is not safely synchronized

Status: open
Status: done

Problem:

• The stats map is mutated and read across goroutines without full synchronization.

9. Unbounded response reads can cause avoidable memory blowups

Status: open
Status: done

Problem:

• Some response paths use unbounded reads from remote servers.

10. File token store has no locking

Status: open
Status: done

Problem:

• Multiple processes can corrupt token storage.

11. syncer.go is a monolith

Status: open
Status: done

Problem:

• One file currently owns protocol setup, planning, batching, auth, stats, and execution.

12. Entry points duplicate setup work

Status: open
Status: done

Problem:

• Run, Bootstrap, Probe, and Fetch all repeat protocol validation, stats setup, connection setup, and ref discovery.

13. Functions carry too much ambient state

Status: open
Status: partial

Problem:

• Large helpers like batched bootstrap depend on too many parameters and too much shared context.

Rewrite requirement:

• Introduce explicit session/context objects with narrow responsibilities.

Performance And Scalability

14. Batch planning probes by repeatedly fetching full packs and discarding them

Status: open
Status: partial

Problem:

• sourcePackExceedsLimit can turn checkpoint planning into repeated full-pack fetches.

15. Materialized fallback path does not scale to large repos

Status: open
Status: partial

Problem:

• The non-relay path stores fetched objects in memory.

16. Fast-forward checks can degenerate into full graph walks

Status: open
Status: done

Problem:

• reachesCommitHash can traverse very large histories.

17. Packet parsing allocates too aggressively

Status: open

Test Gaps

18. Core planning functions are under-tested directly

Status: open
Status: done

Missing direct tests include:

• buildDesiredRefs

19. Relay eligibility logic is only tested indirectly

Status: open
Status: partial

Missing direct tests include:

• canIncrementalRelay

20. Protocol v2 error handling is under-tested

Status: open
Status: done

Missing test coverage includes:

• malformed pkt-lines

21. Missing behavioral coverage

Status: open
Status: partial

Known missing cases:

• empty source repo

22. No benchmark coverage for the expensive paths

Status: open
Status: done

Rewrite Branch Acceptance Criteria

• All mapping validation happens before network activity.
• Capability negotiation is centralized and enforced consistently.
• Relay strategies are separate packages with explicit inputs and outputs.
• Tag creation is correct whether or not a pack transfer is needed.
• Stats and logging are concurrency-safe.
• Protocol parsing has explicit malformed-input tests.
• Rewrite passes go test ./... and go test -race ./....
• Rewrite includes benchmarks for the critical planning and execution paths.
• Rewrite branch can be compared against current behavior using the same integration scenarios.

Suggested Execution Order

7. Rebuild auth and token store behavior.
8. Port and expand tests.
9. Compare rewrite branch behavior against current integration fixtures.