From: Robert_Grant.LAX1B@xerox.com Subject: TECH: Object Interaction Date: Fri, 19 Feb 1993 16:08:05 PST Hi everyone, I've got another question for you all! :-) Has anyone done any thinking about controlling objects? What I mean is, say you become a telephone (why, I don't know!) or a bird, how would you go about mapping the user's flailing about to dialing a number or flying? In my application, I'm passing user events to the object and letting it work out how it wants to process them. However, I get into a bit of a fix when I'm controlling a hierarchy of objects in which some (at different levels) may respond to the same stimulus/event - say the key being pressed. Should I only allow one object to respond to the event? Or all of them? Or do I need to address events (kind of like a nerve net - i.e. send this stimulus to this object)? But then who supples the address? The user or the parent object? It probably should be the parent object as the user does not want to know about the constituent parts of the object that it's controlling - so we come full circle back to mapping events to objects! Can anyone help me out of this cycle? What about sub-objects registering the stimuli they respond to with their parents, do you think that would work? While you're thinking about that - how about this.... How do you gain control of an object? Do you just have to "touch" it? Do you just have to point at it? Do you just have to know it's name? Probably all three! :-) How about the possibility of being able to carry useful stuff around with you, like a "filofax"? As always, any insights are greatly appreciated! Thanks to everyone who's responded to my previous questions! Robert RGrant.lax1b@xerox.com [Mod (Mark): This bears much resemblance to William Bricken's idea of the Virtual Body. People at the HIT Lab have been toying with mapping physical effectors to virtual ones-- perhaps try contacting Max Minkoff, max@hitl.washington.edu.]