I agree in that glass is never 100% transparent. The problem I found with reflection is If I set the transparency to 0.2 the overall end result alpha channel will be multiplied by that value.This will seems to multiply with 0.2 even those fragments with a very strong specular/highlight reflection.(1) Reality: (a) Strong reflection in glass makes objects behind the glass invisible(b) Normal glass without any reflections, adds a little bit of it's tint, depending on thickness, to the color that is visible behind.Compare an image of a pair of glasses with plexi glass side-protectors from my desk: https://imgur.com/cA5ylB0Where there is a reflection the glas becomes basically opaque.(2) SceneKit:(a) High Opacity: Reflections are beautifully reflecting lights etc, unfortunately one can barely see the object behind the glass.(b) Low Opacity: Great visibility of objects behind the glass, but now all the reflections are very subdued.So, what I ended up doing with this is adjusting the alpha-channel of a fragment by the amount of specular lighing that fragment gets:vec3 light = _lightingContribution.specular; float alpha = reflectivity * min(1.0, 0.33 * light.r + 0.33 * light.g + 0.33 * light.b); _output.color.rgb *= min(1.0, (1.5 + 2 * \(minAlpha)) * alpha); _output.color.a = (0.75 + 0.25 * \(minAlpha)) * alpha;`(We put that shader modifier into our release of https://apps.apple.com/app/id1463380262 in case you want to try yourself)@gchiste I also implemented your tip using `blendMode = .add`, which is a good alternative to the above custom shader in same cases.I agree though that it doesn't work perfectly under varying lighting conditions.

If I comment out the .binaryTarget and only use the .target(name:) the bash output is ARCHIVE SUCCEEDED [58.086 sec] but then if I look into the archived file, it has no *.framework inside it.

Mhm, so I thought according to the Swift package Manager Team an Xcode project is not necessary to build an xcframework based on a Swift Package. - https://github.com/apple/swift-package-manager/pull/2981#issuecomment-710282803. boris is also referencing the same demo example he did in the WWDC video Distribute binary frameworks as Swift packages - https://developer.apple.com/videos/play/wwdc2020/10147/ from 2020, so I thought this is the latest status compared to the video Binary Frameworks in Swift from 2019. As a swift package, my project automatically gets an extension that provides the Bundle.module extension, e.g. swift let string = NSLocalizedString(self, tableName: nil, bundle: .module, value: "", comment: "") which is no longer available if I wrap the code into an xcodeproj. I thought based on the above mentioned github issue, the xcodeproj way is no longer necessary.

So it seems just putting the line token preliminary:anchoring:type = "face" in was not enough. What worked was writing a python script, importing the from pxr import Usd, Sdf, UsdGeom, Kind stuff and then creating a scene hierarchy with /Root/Scenes/Scene/Children/MyModel where the scene gets the anchor-token. The USD Classes reference (https://graphics.pixar.com/usd/docs/api/class_usd_stage.html) helped a little but It was way more complicated then I expected it to be.

Seeing the same issue here. Our test scene used to render at 60fps back in April 2021, now without changing any code it's at 2FPS and my whole Macbook Air freezes. Test scene: https://test.looc.io/forest/index.html

We're seeing the same issue in all our WebGL applications. It definitely worked in April 2021, since then without changing any code the Safari performance went from 60fps to 2fps. Example scene built in unity (with a camera-texture since camera input is important for our AR applications): https://test.looc.io/forest/index.html

Turns out homebrew seems to have spammed hundreds of lines eval "$(/opt/homebrew/bin/brew shellenv)" into my ~/.zprofile which I didn't even know existed.

So, now that we have an iPad 5th Generation with the 12MP camera I can see the problem more clearly. Basically, what we'd need is for the ARSCNView to also have the center stage function that Apple offers in FaceTime. The full wide-angle-camera image is too big.

PS: If we remove one of the fields from Resource , say delete the authToken field, the method will no longer crash. I believe that would reduce the size of the struct Resource to where the init method would no longer be outlined

PS: If I change enum HTTPMethod<Payload> {     case get     case post(Payload)     case patch(Payload) } to enum HTTPMethod<Payload> {     case get(Payload)     case post(Payload)     case patch(Payload) } the problem also goes away. Filed a bug report using the Feedback Assistent, there shouldn't be reason why the .get case has an associated value.

I think it would help to specify more clearly what shader system you are using? When it comes to ARKit, ARKit is only the framework that matches your camera to the 3D scene, i.e. handling of the odometry etc. The materials are rendered in the 3D shaders used, technically you can use ARKit with shaders from RealityKit, SceneKit and MetalKit. here's a nice writeup on their differences on stackoverflow

PS: Your title says usdz but you tagged it ARKit. Are you trying to show your object inside an app that uses ARKit? Or are you trying to show a usdz using QuickLook on a website ?