Skip to content

MODERNIZATION-ROADMAP.md

Latest commit

 

History

History
86 lines (64 loc) · 4.48 KB

MODERNIZATION-ROADMAP.md

File metadata and controls

86 lines (64 loc) · 4.48 KB
Raw

DirectFB 1.x Modernization Roadmap (2026+)

Maintainer: Denis Oliver Kropp (deniskropp)
Repository: https://github.com/deniskropp/DirectFB
Vision: Evolve the canonical low-overhead embedded graphics library into a reliable foundation for AI-augmented real-time vision and HMI systems, while preserving its original minimalism and fbdev strengths.

Strategic Context

  • DirectFB 1.x (this repository) remains the authoritative source for the original architecture.
  • DirectFB2 exists as a community revival focused on continued embedded relevance.
  • Your 25+ years of stewardship positions this work uniquely at the intersection of legacy embedded graphics and modern AI/computer vision.

Vision

Protect the original philosophy (minimal resource usage, hardware acceleration, deterministic behavior) while adding the capabilities needed for 2026+ embedded AI pipelines:

  • Efficient rendering of inference results
  • Real-time overlays and HUDs
  • Clean integration with modern inference engines
  • Strong support for DRM/KMS on current hardware

Phased Roadmap

Phase 0 — Foundation Revival (Immediate, 0–4 weeks)

  • Update README with current status, DirectFB2 relationship, and authorship credits.
  • Add GitHub Actions CI for build validation on multiple architectures.
  • Regenerate and publish API documentation.
  • Create clear CONTRIBUTING.md and licensing notes.
  • Deliverable: Modern, welcoming landing experience for the repository.

Phase 1 — Build & Tooling Modernization (4–10 weeks)

  • Make CMake the primary/recommended build system (keep autotools for compatibility).
  • Update compiler standards and address recent C++ header changes.
  • Add static analysis, sanitizers, and basic embedded-focused testing.
  • Improve cross-compilation documentation and scripts for common targets (ARM, RISC-V, NVIDIA Jetson, etc.).

Phase 2 — Backend Evolution (High Priority)

  • Primary Goal: Add or significantly improve a DRM/KMS system module alongside the existing fbdev backend.
  • Optional: Lightweight EGL / OpenGL ES translation layer for capable hardware.
  • Keep fbdev as the pure, ultra-low-overhead path for constrained devices.
  • Align directionally with DirectFB2 efforts (complementary positioning).

Phase 3 — AI & Computer Vision Integration Layer (Core Opportunity)

  • Design a clean AI Surface Import extension for efficient (ideally zero-copy) transfer of inference output tensors into DirectFB surfaces.
  • Provide optimized primitives for real-time overlay / HUD composition (perfect for defect visualization, monitoring dashboards).
  • Support headless/offscreen rendering mode.
  • Define performance contracts (predictable latency, bounded memory).
  • Direct synergy: Ideal companion to autonomous underbody inspection, robotics vision, and other real-time AI pipelines running on embedded Linux.

Phase 4 — Safety, Testing & Packaging

  • Modern test harness focused on embedded constraints and determinism.
  • Targeted memory safety improvements in non-hot paths.
  • Updated packaging recipes (Alpine, Yocto, Buildroot).
  • Clear documentation of Fusion multi-application model on modern kernels.

Phase 5 — Strategic Positioning & Ecosystem

  • Publish positioning statement clarifying relationship between this 1.x maintenance repo and the DirectFB2 revival.
  • Create portfolio / case-study artifact highlighting 25+ years of embedded graphics expertise bridging into AI-hybrid systems.
  • Explore light collaboration or advisory opportunities with the DirectFB2 project if aligned.

Prioritization

High: Phase 0 + Phase 2 (DRM/KMS) + Phase 3 (AI integration) — maximum portfolio and technical impact. Medium: Phase 1 tooling modernization. Ongoing: Phase 4 & 5.

Risks & Mitigations

  • ABI stability: New features behind opt-in flags or new interfaces.
  • Hardware driver maintenance: Focus effort on generic + DRM/KMS paths.
  • Scope control: Every change must map back to an existing or new TAS block.

Success Metrics

  • Clean build + basic examples running on a modern embedded board.
  • Measurable low-overhead rendering path for AI result visualization.
  • Clear, citable narrative of unique maintainer position.

Next Steps

This roadmap was generated as part of ongoing meta-infrastructure work. Contributions, discussions, and prioritization feedback are welcome via issues or pull requests.

Maintained by: Denis Oliver Kropp License: LGPL-2.0-or-later

This document lives in the repository to serve as living guidance for the evolution of DirectFB.