Skip to content

Documentation/readiness: Prioritize the #72, #82, and #101-#123 backlog #124

Description

@smiggleworth

Type

Documentation/readiness

Affected subsystem

Repository-wide open backlog: persisted row encoding, schema concurrency, referential integrity, database images, query execution, specialized indexes, resource controls, pgwire compatibility, and authorization design.

User/operator impact

The newly reported defects mix critical persisted-data risks, silent query errors, bounded compatibility gaps, and one future authorization enhancement. Without a durable dependency order, lower-impact edge work can displace data-integrity fixes and related root causes can be patched inconsistently.

Current behavior

The intake contained 25 open issues: reopened #72 and #82 plus #101-#123. Current source is 8b9716f; the only changes after the reports' 477556c baseline are Admin UI and readiness-document updates, so the cited Rust paths remain current. #109 is a true duplicate of #101 because #101 already requires both the column and expression collectors to use one bound-intersection fix.

No open pull requests or milestones exist. Repository labels provide category labels but no priority labels, so this tracker and each child issue's Priority and dependencies section are the canonical ordering record.

Expected behavior/outcome

Work correctness before optimization and work each root-cause cluster in dependency order. Independent clusters may proceed independently, but a child must not bypass a dependency named below. Every implementation starts with a focused failing regression and closes only after the repository's required validation ladder and exact acceptance criteria pass.

Priority and dependency order

P0 — release-blocking data integrity and recovery

P1 — silent query errors, fail-closed behavior, and resource bounds

P2 — bounded compatibility, adaptive planning, and lower-confidence races

P3 — dependency-bound future hardening

Deterministic evidence

Confidence and support status

High confidence in the metadata and dependency triage. Reports explicitly based only on static source analysis still require their stated red reproduction before implementation claims are accepted. Priority is impact-based, not a claim that every reported runtime symptom has already been reproduced.

Acceptance criteria

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

    documentationImprovements or additions to documentation

    Type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions