The README carried explanations that belong in AGENTS.md; it now follows
the shape of the sibling node repositories and is a quarter shorter.
The tick description was also wrong: it does not trigger a grab. Capture
is free-running at ZED_FPS and a tick forwards the newest frame there is,
so the tick sets the publish rate, and frames captured between two ticks
are dropped. The usage snippet listed four of the five outputs, which
read as an omission rather than the choice it is.
dora-image-view takes std/media/v1/Image, and dora's compatibility graph
has no edge from std/core/v1/UInt8 to it, so wiring the two failed
`dora validate` whatever the node actually sent. Its flat-buffer fallback
would not have decoded these frames either: it reads width/height
metadata, and this node only published shape.
image and depth now go out as the Image struct, camera_info and
point_cloud as Float32 arrays and camera_state as a string, all built
against the Arrow C data interface in src/arrow so the node still links
no Arrow C++. Shape, dtype and encoding metadata are dropped where the
payload describes itself; intrinsics, frame_id and both timestamps are
unchanged. tests/run.sh covers the packing without a camera or the SDK,
round-tripping each array through pyarrow and rebuilding it under the
sanitizers.
Breaking: image_bgr is renamed to image and every payload changes shape.
dataflow_plot.yml is replaced by dataflow_view.yml, which wires the
viewer from a tag and repeats its contract in input_types so validate
checks both sides of the wiring.
Dockerfile provisions CUDA 12.8, ZED SDK 5.2, ONNX Runtime and Rust on
Ubuntu 24.04, the newest release Stereolabs ships a ZED SDK for, so the
node can be compiled and the build verified without installing the SDK
stack on the host.
The compile needs --gpus all: libsl_zed.so links against the CUDA driver
API, NVDEC and NVENC, which ship with the display driver rather than the
CUDA toolkit and so exist in no image. The container runtime mounts them
from the host, and only includes NVDEC/NVENC when the video capability
is requested.
Runs as the invoking user so build/ and bin/ are not left root-owned,
which needs the toolchain outside /root and the ZED SDK directory
readable to an arbitrary uid.
The dora bindings were always built with the debug profile and linked
from target/debug, so a release build still pulled in unoptimized arrow,
zenoh and opentelemetry code -- a 640 MB dora_zed against 48 MB once the
profile follows the build type.
Cloning main made the build non-reproducible and free to drift from the
dora runtime that launches the node. DORA_GIT_TAG can still pin a tag
explicitly, and an installed dora CLI continues to take precedence so
the bindings match the runtime, now with a warning when that CLI
predates the 1.0 this node targets.
The release tarball ships its CMake package files under lib/, but
onnxruntimeTargets.cmake resolves the library as <prefix>/lib64. Ubuntu
is multiarch, so CMake does not search lib64 for packages and ld.so does
not search /usr/local/lib64 at all, leaving find_package(onnxruntime)
failing on a stock install. Keep the libraries in lib64, move only the
package files to lib/cmake, and add lib64 to the linker path.
ONNX Runtime is a configure-time dependency for every build, not only
for the yolo_onnx engine. The event loop no longer inspects the input
id, so it grabs on any input rather than specifically on tick.
Deletes the yoloworld_onnx engine, the CLIP BPE tokenizer, the standalone
test binary and both export scripts. Drops the update_vocabulary input and
the yoloworld_onnx/open_vocab dispatch from main.cc, along with the now
unused Engine::set_labels. Documentation follows.
cristhian