Skip to content

ROADMAP.md

Latest commit

 

History

History
83 lines (63 loc) · 3.49 KB

ROADMAP.md

File metadata and controls

83 lines (63 loc) · 3.49 KB

Fxpmath Roadmap

This roadmap communicates high-level priorities and milestones for upcoming releases. It is intended for contributors and users who want visibility into project direction.

Current Focus: v0.5.0

v0.5.0 is the next major release and is focused on three themes:

  • New capabilities (NumPy interoperability and feature expansion)
  • Better internal structure and maintainability
  • Measurable runtime performance improvements

Milestones for v0.5.0

M1 - Stability and Compatibility Baseline

  • Finalize NumPy compatibility behavior for supported versions (1.26.x and latest 2.x)
  • Resolve remaining correctness edge cases in arithmetic and dispatch paths
  • Add missing regression coverage for known high-risk scenarios
  • Align core NumPy protocol hooks with modern expectations (__array__, __array_wrap__, out behavior)

Exit criteria

  • Full test suite green on required compatibility matrix
  • No known critical correctness regressions open for v0.5.0

M2 - CI and Release Guardrails

  • Enforce required CI checks for lint + tests across supported Python and OS versions
  • Add required NumPy version matrix jobs
  • Add package build/install verification in CI
  • Add release checklist gate for compatibility and quality reports

Exit criteria

  • Required checks are stable and blocking for v0.5.0 branches/PRs
  • Release candidate can be validated end-to-end from CI artifacts

M3 - Core Refactor and Maintainability

  • Reduce complexity in core modules (starting with objects.py) via internal decomposition
  • Replace broad exception handling with typed/explicit error paths where practical
  • Improve test organization and naming consistency
  • Establish lint/style ratchet to prevent maintenance debt growth

Exit criteria

  • Core refactor lands without behavior regressions
  • Maintainability metrics trend positively (complexity/lint debt does not increase)

M4 - Feature Expansion

  • Expand NumPy dispatch coverage for high-value operations (vecdot, matvec, vecmat, norms)
  • Improve behavior consistency for out/out_like and dispatch kwargs
  • Document feature behavior and compatibility caveats

Exit criteria

  • Newly added dispatch features have tests and documentation
  • Public behavior is consistent with documented expectations

M5 - Performance Milestone

  • Add repeatable benchmark suite for arithmetic hot paths (add, mul, truediv, pow)
  • Record baseline metrics for v0.5.0 development
  • Implement targeted optimizations and track deltas
  • Add regression guardrails for performance in CI or release validation

Exit criteria

  • Benchmarks are reproducible and versioned
  • Targeted hot paths show measurable, validated improvements

Release Readiness (v0.5.0)

v0.5.0 is considered release-ready when:

  • Stability milestones (M1, M2) are complete
  • Refactor changes (M3) are merged with no major regressions
  • Selected feature scope (M4) is complete with tests/docs
  • Performance milestone (M5) provides baseline + improvement evidence

Looking Ahead (Post v0.5.0)

After v0.5.0, roadmap priorities are expected to include:

  • Broader API/NumPy feature coverage
  • Additional performance optimization passes
  • Documentation depth and developer experience improvements toward v1.0

Contribution Notes

  • Roadmap items may be adjusted based on bug reports, contributor bandwidth, and ecosystem changes.
  • If you want to help, open or comment on an issue and reference the milestone (M1-M5).
  • Small, focused pull requests aligned to roadmap milestones are preferred.
⚡