Improvements made to sesinetd auto install feature.

Fixed a bug with the helps launch feature trying to launch with the wrong file.

Deleted the FillData function from the generic HAPI C# bindings, specifically HECSharp_StructMembers.cs because it is specific to HE for Unity's implementation of curves.

Fixed a possible crash in IPR when the AOV specified by Variance Pixel Oracle's "Plane" parameter no longer exists upon render restart.

Fixed a bug on the reverse foot where the controls can flip when using negative scales on the joints.

XPU: Fixed a crash when a material update in a scene with nested instances caused Optix device to crash.

setParmsFromData errors because of wrong function args in nested function -Fix a bug, where this function would error when running because of missing

required arguments.

When husk was rendering only non-raster products (i.e. rendering only a deep image with no beauty image), only the first AOV from the non-raster render product was created. This meant that the AOVs might not be available to the render delegate for the non-raster product. Now, husk will create all the AOVs when it creates the dummy raster product.

Added parameters to the Python Processor that allow it to filter and ignore upstream failed dependencies. These options were already available via the pdg.NodeOptions configuration object in Python, but previously weren't promoted up to parameters on the Python Processor itself.

XPU: added new envvar "KARMA_XPU_DEVICES" which allows to enable/disable multiple devices via a single envvar entry (see envvar docs for details). (please note this superceeds the previous KARMA_XPU_DISABLE_DEVICE_n style, but does not replace KARMA_XPU_DISABLE_OPTIX_DEVICE or KARMA_XPU_DISABLE_EMBREE_DEVICE given they have their own benefit of initializing the device type to start with)

Added support for the status operation to the Perforce TOP.

• Added support for Starting/Stopping HAPI Performance monitoring.This can be done by using the Houdini.StartHAPIPerformanceMonitorand Houdini.StopHAPIPerformanceMonitor commands respectively.The performance logs can then be easily viewed when using session syncby opening the Performance Monitor in Houdini.The HAPI traces are stored in the current unreal trace directory.
• Added support for the "unreal_pdg_asset" detail attribute.It can be used to indicate to the plugin if an HDA is expected to use PDG.This is especially useful on big HDAs that contain a lot of nodes - as itwill speed up instantiation and rebuild times significantly.
• Added other optimizations for large HDAs.
• Rebuilding an HDA no longer causes every parameter to be individuallyre-set on the Houdini side.
• The plugin now relies on parameter presets for setting parameter onrebuilt/loaded HDAs. This significantly reduces the time needed for theseoperations on big HDAs that have a lot of nodes and parameters.

• Added support for Houdini Education and Houdini Engine Eduation licenses.
• The "non-bypassed" node flag is now properly considered as recursiveby composeChildList: the function will no longer recurse inside bypassednodes.

Flat camera foreground images now render in Vulkan.

Houdini now includes the OpenCL driver version string when computing the hask key for cached OpenCL binaries, so driver updates will require a re-compile of OpenCL kernels. The driver version is also now output in the About Houdini box and hgpuinfo -c output.

XPU: a slight optimization to pyro volume shaders

• Fixed potential crash when switching render delegate in VK viewport,caused by stale reference left in descriptor sets
• Improved performance in some pick ID render passes

• Reduce number of layers used in High/Medium transparency passes in VK,to reduce memory usage
• Fix instanced primitives not drawing, depending on viewer settings

• Add more detail to log message when Vulkan instance creation fails
• Fix warning messages on startup, when using shader validation

Fixed the documentation for hou.removeContextOption.

Reworded the prompts in the Dynamic Motion tool to be more succinct.

mix() is properly syntax-colored as a built-in function in OpenCL contexts.

The mpmsolver SOP will no longer error out on colliders containing shear transformations. A warning will be displayed to the user instead and the shear component will be ignored.

• Added a new plugin settings "Enable Editable Node Tools",that can be used to change the default value used for itscorresponding property when instantiating HDA.

• Added shared memory session option.

Fixed a bug where PDG/TOPs failed to retrieve logs from tasks submitted to Deadline.