Skip to content

Roadmap

QuicD development roadmap showing planned features, improvements, and future direction.

QuicD is actively developed and production-ready for HTTP/3 and custom protocol use cases.

Stable:

  • ✅ QUIC v1 (RFC 9000) transport
  • ✅ HTTP/3 (RFC 9114) server
  • ✅ Custom application interface
  • ✅ Multi-threaded worker architecture
  • ✅ io_uring zero-copy I/O
  • ✅ eBPF connection routing
  • ✅ OpenTelemetry observability

In Development:

  • 🚧 Media over QUIC (MOQ)
  • 🚧 Connection migration support
  • 🚧 QUIC datagrams
  • 🚧 Advanced congestion control (BBR)

Full implementation of Media over QUIC for low-latency live streaming.

Status: Design phase, basic framework in quicd-moq crate

Features:

  • Track-based publish/subscribe model
  • Multiple delivery modes (streaming, datagram, fetch)
  • Integration with media pipelines (FFmpeg, GStreamer)
  • Low-latency streaming (< 1 second glass-to-glass)

Timeline: Q1 2025

Use cases: Live streaming, video conferencing, gaming

Complete QUIC connection migration support.

Current: Basic CID management implemented

Planned:

  • Active path validation
  • Automatic failover on path failure
  • Multi-path support
  • NAT rebinding detection

Timeline: Q2 2025

Use cases: Mobile clients, network changes, WiFi to cellular handoff

Full support for unreliable datagrams (RFC 9221).

Current: Partial implementation

Planned:

  • Send/receive datagram API
  • Flow control integration
  • Application-level datagram framing

Timeline: Q1 2025

Use cases: Gaming, VoIP, real-time telemetry

Pluggable congestion control with BBR and custom algorithms.

Current: CUBIC only

Planned:

  • BBR v2 implementation
  • Pluggable congestion control API
  • Per-application congestion control selection
  • Congestion control tuning UI

Timeline: Q2-Q3 2025

WebTransport server for browser clients.

Planned:

  • WebTransport protocol implementation
  • Browser-compatible API
  • Certificate management for dev/prod
  • Integration with existing HTTP/3 server

Timeline: Q3 2025

Use cases: Browser-based applications, WebRTC alternatives

Continuous performance improvements.

Planned:

  • DPDK integration for kernel bypass
  • Hardware offload (crypto, checksumming)
  • Improved memory allocator
  • Better cache locality
  • Profile-guided optimization (PGO)

Timeline: Ongoing

Targets: 40+ Gbps per instance, < 10μs tail latency

Extended telemetry and debugging capabilities.

Planned:

  • Distributed tracing with OpenTelemetry
  • eBPF-based profiling integration
  • Per-connection performance metrics
  • Live debugging interface
  • Performance regression detection

Timeline: Q2-Q3 2025

Unified server for multiple protocols.

Planned:

  • HTTP/2 over QUIC
  • DNS over QUIC (DoQ)
  • RTP over QUIC
  • Custom protocol templates
  • Protocol negotiation framework

Timeline: 2026

Native load balancing and clustering support.

Planned:

  • Connection migration for load balancing
  • Consistent hashing for routing
  • Health checking and failover
  • State synchronization
  • Anycast support

Timeline: 2026

Advanced security features.

Planned:

  • Post-quantum cryptography
  • Certificate transparency
  • Advanced DDoS mitigation
  • Rate limiting framework
  • Security audit mode

Timeline: 2026+

Improved developer tools and experience.

Planned:

  • QuicD CLI tool with scaffolding
  • Hot-reload for development
  • Built-in testing framework
  • Protocol simulation/fuzzing
  • Visual debugging tools
  • IDE integrations

Timeline: Ongoing

Vote on features or request new ones:

Based on community feedback:

  1. WebTransport (12 votes) - In planning
  2. Windows Support (8 votes) - Blocked by io_uring dependency
  3. HTTP/2 Compatibility (7 votes) - Planned for 2026
  4. Docker Support (6 votes) - Planned for Q1 2025
  5. Connection Pooling (5 votes) - Under consideration
  • HTTP/3 server implementation
  • Multi-threaded worker architecture
  • io_uring integration
  • eBPF routing
  • OpenTelemetry metrics
  • Custom application interface (quicd-x)
  • Event-driven connection handling
  • Stream management API
  • Initial QUIC v1 transport
  • Basic connection handling
  • Single-threaded architecture

We welcome input on the roadmap! Here’s how to contribute:

  1. Discuss: Comment on GitHub Discussions
  2. Vote: React to issues with 👍
  3. Propose: Open enhancement issues
  4. Implement: Submit PRs for planned features

See Contributing Guide for details.

QuicD follows Semantic Versioning (SemVer):

  • Major (1.0.0): Breaking API changes
  • Minor (0.x.0): New features, backward compatible
  • Patch (0.0.x): Bug fixes

We aim for:

  • Minor releases every 3 months
  • Patch releases as needed
  • Major releases when API is stable (v1.0 planned for 2025)

APIs that won’t change without major version bump:

  • QuicAppFactory trait
  • ConnectionHandle core methods
  • Configuration file format
  • QUIC protocol compliance

Features that may change in minor versions:

  • MOQ implementation (marked experimental)
  • Internal APIs (not part of public API)
  • Performance tuning parameters
  • Telemetry metric names
  • Features marked deprecated in release N
  • Deprecated features removed in release N+2
  • Minimum 6 months between deprecation and removal
  • Clear migration path documented
  • Linux 5.1+: Full support (io_uring required)
  • x86_64: Primary platform
  • ARM64: Tested and supported
  • 🚧 macOS: Planned (will use kqueue instead of io_uring)
  • 🚧 Windows: Under consideration (io_uring alternative needed)
  • 🚧 FreeBSD: Community request
  • x86_64: Fully tested
  • ARM64: Supported and tested
  • 🔄 RISC-V: Experimental
  • Feature Releases: Every 3 months
  • Patch Releases: As needed for bugs/security
  • LTS Releases: Planned starting with v1.0

Last updated: December 2024