Voice SDK Dictation Not Working in Unity 2023.2+
I upgraded my project to Unity 6 from Unity 2022 LTS. Here's my current setup: Unity 6000.0.27f1 Meta XR Core SDK 69.0.1 Meta XR Interaction SDK Essentials 69.0.1 Meta XR Voice SDK 69.0.0 With this setup, the Core SDK requires the use of GameActivity, but the Voice SDK's dictation functionality requires Activity. As a result, dictation cannot be executed on Unity 2023.2+. This is an error on Quest 3. Is this a known issue? Or is there a possible workaround? Here’s a related information for reference: https://communityforums.atmeta.com/t5/Unity-Development/Important-Unity-6-information/td-p/1252153 https://discussions.unity.com/t/voice-sdk-crash-on-activate-meta-quest-3/935246
[RESOLVED] Building Blocks (v65): Controller Tracking visuals for Quest3 are not wired correctly
Hi, after using Building Blocks in Unity to set up controller tracking I realized, that the visuals for the Quest Plus controllers are not wired up correctly. Pls see the attached screenshots. It's simple to solve, but annoying to do manually every time. 😑 br, Max Problem - Left Controller: Problem - Right Controller: Solution - Left Controller: Solution - Right Controller:
Consistent Unity Playmode Runtime Error when using Link after updating the Quest Link app
I updated my meta quest link app to version 71.0.0.167.524 on the date 9/30/2024. Since I have updated earlier this week, I have a new error that I face everytime I click into play mode in my editor with link. The error is "Microsoft Visual C++ Runtime Library Runtime Error! Program: This application has requested the Runtime to terminate it in an unusual way. Please contact the application's support team for more information." I tested, around the error and it does NOT occur when running play mode normally or with the meta XR simulator, and it DOES occur 10/10 times when starting playmode with Quest Link open, which confirms the culprit. I searched the internet and the unity logs for more information but wasn't able to find answers. Please let me know if there are more logs that would be useful to attach from the Quest Link software. This error is blocking my work everyday this week, i'm working on a game that is about to ship, and now I cant get in headset to test in editor anymore. Its a shame I was just complementing folks at a meta about how much better link has gotten for developing mixed reality apps. The next things I will try are fresh installs of the project, fresh installs of the editor, then maybe a fresh install of windows. I hope this post gets noticed by others with this issue or someone at meta who can help fix it. Fingers crossed ill get another update any day here and I can get back to work. Versions and Info Windows 11 PC Quest 3 Connected via USB-C Unity Version:2022.3.20f1 Quest Link Version: 71.0.0.167.524 Quest Headset Version: 69.0.0.576.352.641228209
Enabling Meta XR Feature Group in OpenXR breaks controller tracking
Using Unity 2022.3.4f1 LTS with the updated packages for OpenXR (1.8.1) and Oculus XR Plugin (4.0.0) and v54.1 of the Oculus Integration SDK, enabling the `Meta XR feature group` under XR Plug-in Management/OpenXR completely stops the controllers from tracking. The controllers are not visible in the Analysis/XR Interaction Debugger/Input Devices and no device connection events are raised from InputDevices. Disabling the Meta XR feature group fixes the behaviour. I have tried earlier versions of the integration SDK but anything < v53 yields a compile error with the new OpenXR and Oculus plugins. Is this a known issue?
NavMesh not recognised in IL2CPP builds
We have recently run into some problems with navigation in a Quest 2/3 project. Navmesh agents which functioned perfectly in the editor were not working in our test builds for android. After some testing we've learned that when building for Windows or Android, if the Mono backend is available and used, everything behaves as expected, but switching to IL2CPP causes the below error messages, regardless of whether the navmesh agent is pre-existing or instantiated at runtime. I have already tried many potential solutions including Warp() or setting position before enabling the agent, rebuilding NavMeshSurfaces at runtime before instantiating the agents, and have switched back to the legacy NavMesh system to rely on baked navmesh data, none of which have changed anything. We have tested with extremely basic Navmesh agent wander behaviour and it seems clear that the backend is somehow affecting the references to Navmesh or the mesh data itself. Has anybody experienced this problem, or do you have any ideas for workarounds/solutions? I hope someone can help, thanks!
PointerExit events sent when clicking UI button (PointablecanvasModule.cs) SDK v63.0.0
Found an issue when clicking a UI button with custom OnPointerEnter / OnPointerExit events. Issue: PointerExit events getting triggered when clicking a button even though the pointer never leaves the button. Setup: Meta XR Interaction SDK 63.0.0 PointableCanavsModule (With EventSystem) and a Canvas with a standard Button. The button has a custom script that registers OnPointerEnter and OnPointerExit events, each event is printing a message in the log: "ENTER" and "EXIT". When clicking the button while keeping the pointer on the button (never breaking the hover) you get this log: EXIT EXIT ENTER It looks like the Exit event is hardcoded to be sent whenever a click has happened which I believe is a problem in itself, in addition I don't see why it would fire twice on top of that? As far as I can see, For UI to work you need to add PointableCanvasModule.cs from the Meta XR Interaction SDK. This will call ProcessPointer on each of the pointers which if released (I assume this means "clicked") will call the base class in Unity UGUI package (EventSystem/InputModules/BaseInputModules.cs) using this call HandlePointerExitAndEnter(pointerEventData, null); But it is always passing NULL as newEnterTarget and if null the code below gets executed and looks like it will send exit events no matter what. I would say this is a bug in the Meta XR Interaction SDK / PointableCanvasModule.cs. I cannot imagine this being a feature as it adds quite a bit complexity for using OnPointerEnter and OnPointerExit for custom hover functionality on clickable buttons. BaseInputModules.cs // walk up the tree till a common root between the last entered and the current entered is found // send exit events up to (but not including) the common root. Then send enter events up to // (but not including) the common root. // Send move events before exit, after enter, and on hovered objects when pointer data has changed. protected void HandlePointerExitAndEnter(PointerEventData currentPointerData, GameObject newEnterTarget) { // if we have no target / pointerEnter has been deleted // just send exit events to anything we are tracking // then exit if (newEnterTarget == null || currentPointerData.pointerEnter == null) { var hoveredCount = currentPointerData.hovered.Count; for (var i = 0; i < hoveredCount; ++i) { currentPointerData.fullyExited = true; ExecuteEvents.Execute(currentPointerData.hovered[i], currentPointerData, ExecuteEvents.pointerMoveHandler); ExecuteEvents.Execute(currentPointerData.hovered[i], currentPointerData, ExecuteEvents.pointerExitHandler); } currentPointerData.hovered.Clear(); if (newEnterTarget == null) { currentPointerData.pointerEnter = null; return; } } ... The above might not be the reason for this behaviour, but the behaviour is there nonetheless. Is this something that will be fixed? Pretty sure this was not an issue when using OVRInputModule.cs instead of PointableCanvasModule.cs.Bug Report: Accumulation of Black Passthrough Quads due to OVRTrackedKeyboard
I'm using the Meta All-in-one SDK (v62), and there's a minor bug that is impacting the user experience. Whenever a keyboard is loaded, it creates a passthrough "Quad" with it. The problem arises when it creates a new "Quad" without destroying the existing one, leading to the scene accumulating numerous black passthrough Quads. You can easily reproduce it by toggling "KeyboardQueryFlags" between Local and Remote repeatedly. Its in OVRTrackedKeyboard.cs Script, function : LoadKeyboardMesh() variable : passthroughQuad_Spatial Anchors bug: LoadUnboundAnchors returns null UUIDList
Hi, I am struggling trying to work with Spatial Anchors and what looks like a bug. I am creating a few gameobjects in the virtual room, adding Space Anchors to them, saving the room anchors to local storage and reloading them. I save the uuid list so I can reload the objects even if I close the app and run it again. It works fine but, in some "random" situations, the LoadUnboundAnchors returns a null UUIDList as if it were no Unbound anchors to be loaded. When this happens I can try and try to load them but always get the same result. But... if I create a new anchor in the session and save it, then I can reload all of them.🙄 Once they are loaded I can remove all objects and anchors and reload again and works fine. So, it looks that I cannot load correctly the saved anchors unless I have previously saved one in the same session. I know that they are correctly saved because after creating a new anchor in the empty room I reload them succesfully. Anyone else have this problem? I am using Quest 2, Unity 2022.3.11, Meta XR Core 59. The loading sequence is something like this: OVRSpatialAnchor.LoadOptions options = new OVRSpatialAnchor.LoadOptions { Timeout = 0, StorageLocation = OVRSpace.StorageLocation.Local, Uuids = GetSavedAnchorsUuids() }; OVRSpatialAnchor.LoadUnboundAnchors(options, _anchorSavedUUIDList => { if (_anchorSavedUUIDList == null) { //Debug.Log("Anchor list is null!"); return; } .... }Passthrough and other functionalities broken on v57
So I was updating a project in unity to the v57 and all the passthrough functionalities broke. Im trying the default Passthrough scene and even that one doesnt work. Also the spatial anchors doesn't work neither. Even a new empty project with the oculus integration package doesn't work the example scenes of Passthrough or spatial anchors. Any idea on how to solve it?
