Skip to content

BACKLOG.md

Latest commit

 

History

History
62 lines (55 loc) · 2.85 KB

BACKLOG.md

File metadata and controls

62 lines (55 loc) · 2.85 KB
Raw

Product Backlog

This file records product work requested or accepted as part of the cleanup plan and its delivery status.

Data Integrity

Virtual FK orphan validation

  • Status: implemented
  • Goal: block creation of a virtual FK when the proposed relationship would connect orphan records on either side of the join.
  • Expected behavior:
    • validate real data compatibility before persisting the virtual FK
    • if orphan rows exist, do not create the relationship
    • show an explicit warning explaining why creation was blocked
    • generate a diagnostic SQL query that lists the orphan rows responsible for the block
  • Rationale:
    • keeps the SQL Map surface honest
    • prevents the editor from suggesting a relationship that real data does not support
  • Verification:
    • worker-based bidirectional orphan preflight blocks invalid relationships
    • the blocking dialog exposes diagnostic SQL for the incompatible rows

SQL Map UX

Floating drag affordance for virtual FK creation

  • Status: implemented
  • Goal: when the user drags one field onto another in the SQL Map, the relationship should visually follow the pointer instead of appearing only after drop.
  • Expected behavior:
    • render a floating relationship draft that tracks the cursor
    • preserve readable source/target hover states during the drag
    • keep the effect lightweight and legible rather than ornamental
  • Rationale:
    • improves perceived direct manipulation
    • makes the virtual FK gesture easier to understand before drop
  • Verification:
    • an SVG draft edge follows the pointer and highlights compatible targets
    • the same relationship workflow is available with keyboard focus, Enter/Space, and Escape

QA Tooling

Table population workflow for synthetic data

  • Status: implemented
  • Goal: add a QA-oriented table seeding workflow so users can populate a chosen table with configurable volumes and per-column generation rules.
  • Expected behavior:
    • inspect the selected table schema and list the columns before insertion
    • keep PK fields on the automatic path by default
    • allow per-column strategies based on type, including:
      • fixed numeric values
      • numeric increments with configurable start/step
      • random numeric ranges
      • fixed text
      • random or lorem-style text
      • other type-appropriate strategies introduced later
    • let the user define the target record count
    • insert the requested number of rows into the current database
  • Intended use:
    • QA data setup
    • performance and load experiments against app databases
    • local stress tests such as 100_000 or 1_000_000 row scenarios
  • Verification:
    • table actions open a schema-aware configuration dialog with type-specific strategies
    • insertion runs in one worker transaction with progress, rollback, cancellation, and a 1_000_000 row ceiling
    • volumes above 100_000 rows require explicit confirmation