Open sabraaap opened 1 year ago
The backend of the WebGL template uses some slightly different logic compared to the python controller. PutObject
requires the input of a target sim object, either by objectId
or by some sort of screen based raycast to get the coordinates of the surface of a valid receptacle object.
Because of this, passing in a target receptacle via objectId
is not really an option without a major rework to the web template, and because the web client uses WASD for movement and mouselook for aiming, it is also difficult to guarantee a clear raycast while holding different sized objects, as it was never really designed to be used in that way.
I believe it would be possible to alter the WebGL logic to allow PutObject, but it would be a fairly major rework as the WebGL interface was designed to be just a quick way to navigate and get a feel for the scenes, rather than extensive data collection or experiments.
Hi,
We're keen to reuse the WebGL template for a project but noticed that the PutObject action is not implemented, but only the DropObject action. This is problematic when trying to place objects in receptacles (e.g. an apple in a bowl). Is there a specific reason for this behaviour, or would it be possible to enable the PutObject action?
Thanks!