Date: Mon, 26 Nov 2001 12:18:37 -0500 From: "Chad Wingrave" Subject: interfaces in a 3D environment In-reply-to: <3BFDC956.92ADF0F1@cs.yorku.ca> Sender: To: "3D UI list" <3d-ui@hitl.washington.edu> Message-id: MIME-version: 1.0 X-Mailer: Microsoft Outlook IMO, Build 9.0.2416 (9.0.2910.0) Content-type: text/plain; charset=iso-8859-1 Content-transfer-encoding: 7BIT Importance: Normal X-MSMail-Priority: Normal X-MimeOLE: Produced By Microsoft MimeOLE V4.72.3110.3 X-Authentication-warning: torch.hitl.washington.edu: majordom set sender toowner-3dui@hitl.washington.edu using -f X-Priority: 3 (Normal) We are talking about different subtleties in interaction techniques and those differences are based upon the task, the object, the environment, etc. I think that keeping context with an interface will only help usability because the environment will be able to reason about what the user is doing and do some of the work for them with "magic". Some of the work with Smart Objects is interesting because it lets information about the usage of an object exist in an object. This however is only part of the trouble though since it only captures one behavior of the object at a time. Given that we are dealing with the human perception and cognition which is VERY flexible, I think we should as a community focus less on techniques and more on the Nuances ( <-shameless self promotion ) of all the parts of a VE and how those parts interact. After all, people do not know how they want to work in a VE, they just know how they will respond to affordances and feedback in the VE. This includes the interface hardware and displays. If we spend our time as a community discussing each nuance of every technique for each context, I am afraid we will be much like Monty Python in the discussion of coconut carrying swallows: "Guard1: Listen, in order to maintain air-speed velocity, a swallow needs to beat its wings 43 times every second right?..." "Guard2: It could be carried by an African swallow!" "Guard1: Oh, yeah, an African swallow maybe, but not a European swallow, that's my point." "Guard2: Oh, yeah, I agree with that..." - Monty Python and the Holy Grail Comments? -Chad Wingrave (cwingrav@cc.gatech.edu) http://www.cc.gatech.edu/~cwingrav/ On Thu, 22 Nov 2001, Wolfgang Stuerzlinger wrote: > Jeff Pierce wrote: > > > > At 09:12 AM 11/21/01, Wolfgang Stuerzlinger wrote: > > >Taking this argument one step further one can use what most people percieve > > >as normal or "common sense" (e.g. that gravity exists :-) and use this to > > >simplify interactions even further. This improves interaction speed by > > >a factor of approximately 2.5. See the papers on the MIVE system > > >on my WWW page. > > > > Improves interaction speed for what task? There are some obvious examples > > (manipulation of objects at a distance, navigating large distances) where > > using what people perceive as normal (or, more accurately, familiar) is > > much, much slower than using a "magic" technique. > > True. However, I would like to point out that even "magic" techniques > can benefit from "common sense". Take e.g. a 6 DOF tracker, where > the tracker position & orientation are used to do ray-casting for one > of many interaction-at-a-distance techniques. Users expect e.g. > a distant house to stay flat on the ground when it is moved around, > yet the 6DOF tracker makes this practically impossible, due to rotation > around the ray axis. "Common sense", be it implemented as gravity or > as an "on-floor" constraint, will make a lot of difference even in > this case. > > > And as a minor point, I can posit virtual worlds where you _don't_ want > > gravity. Why on earth would I want things to fall to the floor if I > > accidentally drop them? It'd be much easier if they hovered within reach > > so that I could grab them again. > > I care to disagree. My experience with taking people "off-the-street" > into VR environments is that they don't like objects floating in the air > in essentially arbitrary positions & orientations. One issue here is > that floating objects often block the view of other objects, which > hinders people. Another is that the orientation of the object makes it > often inconvenient to further manipulate the object. More importantly, > people working in zero-g (astronauts) don't leave things floating in > space, even though they could theoretically. They attach objects as much > as possible onto surfaces or other objects to keep things surveyable and > to maximize space for moving around. > > But yes, I do agree that it sounds good to just "hang something into > the air" and there are some interesting uses of this (e.g. Mark Mine's > over-the-shoulder delete/recover idea). I just think that floating > objects is a bad choice for the system-wide default in VR systems/toolkits. > > Wolfgang > -- > Wolfgang Stuerzlinger Dept. of Computer Science; York University > CCB 3048; 4700 Keele Street; Toronto, Ontario, M3J 1P3; Canada > wolfgang@cs.yorku.ca http://www.cs.yorku.ca/~wolfgang >