From: Jeff Pierce [jpierce@cs.cmu.edu]
Sent: Tuesday, August 03, 1999 12:46 PM
To: Robert W. Lindeman
Cc: 3d-ui@hitl.washington.edu
Subject: Re: Idea for discussion: violating assumptions

At 11:16 AM 8/3/99, Robert W. Lindeman wrote:
>Objects have substance
>  - Turn off collision detection (if it was ever turned on).

Here you go: we can extend this into "Objects have weight/mass/substance".
Then you can do things like blow objects around with a little fan.  =)

>To turn things on their head, I would argue that many (most?) interfaces
>routinely violate assumptions. Which brings me to another point. There is
>also the notion of the source of the assumptions. Some come from the
>physical laws, some abstract ones come from the things we have learned
>by using GUIs, and some from other technologies, like 
>telephones/televisions. Anyway, I guess I'm just trying to say that it
>might be important to look at the source of the assumptions.

Very possibly.  I've jotted a few ideas down about origins, but haven't
spent time doing any hard thinking on this yet.


>(I think some of your things from #5 below could fit here too. The thread
>being non-linearity).

I'll definitely eventually need a better hierarchy; many of the assumptions
I listed really play with non-linearity in different ways.


>> 9) The world we perceive really is 3D
>> 
>> * pulling away current view as 2D image to reveal another reality
>> * "pop up" worlds that user can fold up and put away
>
>- Web pages?

I'm not sure exactly what you mean.  Could you say more?


>Others I can think of:
>a) Your avatar looks like you
>   - "On the Internet, no one knows you're a dog."

Good one.  I once toyed with the idea of essentially allowing the user to
shape change in VR.  What if you could be an octopus or an elephant or
(just for Jaron Lanier) a cuttle fish?

>b) You only have 2 hands
>   * Can we activate/manipulate multiple (>2) hands?

Another good one.  You can imagine having hands you're not currently using
holding stuff you'll need later.

I thought about something similar awhile back for CSCW: what if you attach
multiple arms to the user, and then allow remote users to work different
pairs?  The working title was "Octopus Telepresence".  ;)

>c) All objects are oriented alike, wrt the UP vector.
>   - Why is it that all Star Trek ships always meet with the same UP
>     vecotr?

This always amused me about Star Trek.  Star Wars sometimes does a little
better, but they still don't go completely free with orientation.

>Some thoughts...
>1) It might be a good idea to insure reliability, if not predictability.
>What I mean is that an interface should act *consistently* if it is to be
>usable. It can still be fun (and powerful), but it needs to be usable.
>(Unless your goal is pure fun, of course).

No argument here.

>2) It is difficult to tell if you are looking at:
>   a) Navigation,
>   b) Maneuvering,
>   c) Object manipulation,
>   d) General interaction, or
>   e) Fun and Mayhem.

Overall I'm looking at general interaction, although a lot of times I'll
pick a specific task (ex. object manipulation) and then run through the
assumptions to see what I can come up with.

Good discussion!

Jeff