ILLIXR v4.2 Released

ILLIXR v4.2 is out, marking a major expansion of ILLIXR's platform support with native Windows and Android builds alongside Linux. Windows support [#474](https://github.com/ILLIXR/ILLIXR/pull/474) moves Windows dependency management to vcpkg-based source builds, reworks Vulkan initialization for Windows' stricter struct-initialization rules, and updates OS-specific code across the build system and a first wave of ported plugins (native_renderer, offline_cam, offline_imu, passthrough_integrator, rk4_integrator, timewarp_vk, fauxpose, pose_lookup, pose_prediction, vkdemo, tcp_network_backend). A companion cleanup removed all symlinks from the codebase, since Windows doesn't support them [#484](https://github.com/ILLIXR/ILLIXR/pull/484).

On Android, the previously unreleased Android runtime has been merged into the main ILLIXR tree [#485](https://github.com/ILLIXR/ILLIXR/pull/485) with full Gradle/Java build support, and a broad set of plugins now run on-device: a new android_data capture plugin superseding android_imu_cam [#488](https://github.com/ILLIXR/ILLIXR/pull/488), Android builds of gldemo [#489](https://github.com/ILLIXR/ILLIXR/pull/489), offline_cam/offline_imu [#491](https://github.com/ILLIXR/ILLIXR/pull/491), the TCP and UDP network backends [#492](https://github.com/ILLIXR/ILLIXR/pull/492), rk4_integrator and timewarp_gl [#493](https://github.com/ILLIXR/ILLIXR/pull/493), and a GPU-accelerated offload_rendering_client with HEVC and AV1 decode [#494](https://github.com/ILLIXR/ILLIXR/pull/494). A new openxr_interface plugin suite [#490](https://github.com/ILLIXR/ILLIXR/pull/490) relays head/hand poses and hand-interaction data from the Android OpenXR runtime to a remote server over UDP and renders decoded stereo frames directly into the device's OpenXR swapchain, enabling offloaded rendering on Android hardware. Services plugins gained Android-only common_lock and extended_window components [#486](https://github.com/ILLIXR/ILLIXR/pull/486).

Supporting all platforms: a new UDP-based network plugin selectable per-topic alongside TCP [#479](https://github.com/ILLIXR/ILLIXR/pull/479), a latency-tracking plugin for round-trip and one-way client-server measurements on desktop [#481](https://github.com/ILLIXR/ILLIXR/pull/481) and Android [#497](https://github.com/ILLIXR/ILLIXR/pull/497), and a server-side NVENC offload-rendering plugin [#483](https://github.com/ILLIXR/ILLIXR/pull/483) that pairs with the new Android rendering client. Underneath it all, the pose and serialization system has been refactored into a more modular, OpenXR-compliant hierarchy with velocity-aware head poses, enabling prediction from a single pose sample [#480](https://github.com/ILLIXR/ILLIXR/pull/480).

The new server-side `semanticxr_python` plugins [#496](https://github.com/ILLIXR/ILLIXR/pull/496) bridge ILLIXR to [SemanticXR](https://arxiv.org/pdf/2606.12849) (ISMAR 2026, TVCG special issue), a device-cloud system for real-time, open-vocabulary semantic mapping and querying ([demo](https://youtu.be/wNirwDF1wNQ)) on power-constrained XR devices. The plugins hand the incoming RGB, depth, and pose stream to the SemanticXR Python backend and return query results to the client. On Meta Quest 3, ILLIXR streams the sensor data up and renders the returned object point clouds as world-aligned overlays. Mapping and perception code stays in the SemanticXR Python repository.

This release also introduces integration with [Boba](https://jianxiapyh.github.io/Boba-project-page/) (ECCV 2026), a system supporting efficient, power-aware single-instance execution and scalable batched simulation of physics-based Gaussian digital twins, with applications in XR and robotics. The new boba_immersive and boba_streaming_server plugins integrate Boba’s [immersive demo](https://jianxiapyh.github.io/Boba-project-page/#demos) with ILLIXR [#498](https://github.com/ILLIXR/ILLIXR/pull/498). Boba’s simulation and rendering code remains in its own repositories, while ILLIXR delivers Meta Quest 3 headset and controller input to the desktop backend and streams the resulting stereo frames back to the headset.

The full release notes and ILLIXR v4.2 can be found at [https://github.com/ILLIXR/ILLIXR/releases/tag/v4.2.0](https://github.com/ILLIXR/ILLIXR/releases/tag/v4.2.0).