The better way to handle this is to initially have a CloudTracker with an ImageTrackable attached and no augmentations. Then, when OnRecognitionResponse is called, you download the bundle (which shouldn't contain any ImageTrackable objects, just simple augmentations) and add it as a child GameObject to the ImageTrackable, making sure the transform is reset when doing so. The, when another OnRecognitionResponse comes in, you check if the augmentation for that target is already downloaded and added. If not, destroy or disable the existing augmentations and repeat the process.
Would your method allow for more than one augmentation to be present within the scene? So two targets could use the same trackable?
It should be possible. As I've said, when another OnRecognitionResponse comes in, you can check what existing augmentations you already have downloaded and decide if the new target requires a new augmentation, or if it can reuse something already existing, based on the name.
Ok, I think there is a disconnect. What I am trying to do is make the dino sample but with cloud targets rather than the wtc on the device. Will this work? Can I do multiple cloud tracking?
The SDK currently doesn't support multiple (simulatanious) tracking of images in the Cloud.
Is there any road map? I know competitor sdks can do multi image tracking from cloud
I am trying to make the trackable objects into assetbundles, then on OnRecognitionResponse I am downloading the recognized bundle and adding it to the cloud tracker. So this is not working, when adding the trackable to the tracker it is set inactive. When active it does not track. I am trying this approach to try to get multiple cloud trackables to work. Is this even possible or am I going down a rabbit hole?