hi LLT.

unfortunately, phys.multiple creates instances of physics rigid-bodies, not instances of jit.phys.body objects, so you can't use the phys.picker - jit.proxy combination like this.

what you can do, however, is retrieve the index number of the body instance, from the output of phys.picker, and use that to create the matrix cell address, using division and modulo operations.

i modified your patch to better explain my meaning.

please note: there is a bug in the current release of phys.multiple that prevents the mass attribute being set in the max box. we will have an incremental updated posted soon that fixes this.

Hi Robert. Ok, thank you for the patch.

Laurent

heads up: the incremental update is now live, and fixes the issues with setting mass on a phys.multiple. https://cycling74.com/downloads/incremental/

Hi, thanks Robert

hi, am i right that it is not possible to pick a specific instance of jit.gl.multiple with jit.gl.picker, but only with jit.phys.picker via jit.phys.multiple? (Looking for an unique name...)

yes, you are correct. this is a limitation of gl.multiple with gl.picker, but phys.multiple and phys.picker work as expected. you can therefore wrap a gl.multiple with a phys.multiple, and disable dynamics in your phys.world, in order to do picking with phys.picker.