From: owner-3dui@hitl.washington.edu on behalf of Roy Ruddle [Ruddle@Cardiff.ac.uk] Sent: Wednesday, November 29, 2000 12:48 PM To: 3dui List Subject: Re: notation for 3D interaction techniques Nice idea Doug, Here's my tuppence worth: > little thought to realize that the virtual hand should always > remain along a vector starting at the user's torso and going > through the user's physical hand ... probably still ambiguous! yup, neither "torso" or "physical hand" are points so the vector isn't precisely defined > Now for the question: what are the _primitives_ (e.g. 3D cursor, virtual > hand, eye, virtual object), _actions_ (e.g. button press, gesture, > voice input, virtual object selection), and _state_ (e.g. selected > object, virtual hand position) that we would need for a notation for > 3D interaction techniques? Well I'm working on interaction techniques that allow virtual objects to be manipulated in cluttered spaces. Central to the behaviour of those is what happens when a collision occurs, and that can get very complicated. To help, I created flow charts to design the algorithms. Rather than leaping in to define a set of primitves, it might be a good idea to start by writing down the detailed workings of many different types of interaction algorithm in the form of flow charts (or state transition diagrams?). Decomposition of those might lead to primitives for a standard notation, or maybe some other kind of approach (pseudo-code?). The end product should be something that is mathematically precise, and which someone else could use to implement the algorithms in working form. Final thought, what have people's experiences been of using the User Action Notation for 2D stuff? roy