Weekly Discussion Group on Future Computing Environments
Winter Quarter 1997
Alternative User Interfaces
Time and Place: Mondays 4:30-6:00pm,
Room 102, College of Computing
The general focus this quarter is on alternative user interfaces.
By alternative, we mean any interface that moves away from the now
traditional GUI desktop with keyboard/mouse for input and graphic
display as output.
Possible topics to cover this quarter are:
- multimodal interfaces
- immersive environments
- personal vision systems
- the home as a user interface
Blair MacIntyre and Steven Feiner, Future Multimedia User
Interfaces. Multimedia Systems
- On-line meeting notes
No meeting (Martin Luther King holiday)
The Limits of Direct Manipulation
In order to discuss the limits of DM, we need to define it (Hutchin's
paper helps with this) and to think about other types of interfaces
with which we can compare it. Try to come up with one example of
a non-direct manipulation interface, and compare it's abilities and
limits to those of DM interfaces.
From the readings, please pay special attention to Hutchin's
discussion of problems with Direct Manipulation (starts p. 336). I
also found the violin-piano example on p. 328 to be interesting. In
the Bowman & Hodges paper, think about the needs (and limitations) of
3D and VR interfaces with respect to direct manipulation. The
Neurosurgical Planning paper is a good "story" paper for direct
They are be available for photocopying in my office.
- "Assigned" Papers (Bibliography)
- Hutchins, E., Hollan, J., and Norman, D. "Direct Manipulation
Interfaces." in User Centered System Design, D. Norman and S.
Draper, eds., Lawrence Erlbaum Associates, Hillsdale, NJ, 1986,
- Bowman, D. and Hodges, L. "An Evaluation of Techniques for Grabbing and
Manipulating Remote Objects in Immersive Virtual Environments." to appear
in Proceedings of the ACM Symposium on Interactive 3D Graphics, 1997.
- Goble, J., Hinckley, K., Pausch, R., Snell, J., and Kassell, N.
"Two-Handed Spatial Interface Tools for Neurosurgical Planning."
IEEE Computer, vol. 28, no. 7, July 1995, pp. 20-26.
- Pointers to other interesting material
2-D interfaces are not a done deal
In reading these papers, keep in mind the notion of a
"view" in the interface. In the case of paper 1 there are lots of views
presented by the lenses. In the second paper, he talks about the different
"views" that the different card players need in the gin example. Finally, in
the last paper there are many smaller views of the main view. Note that in all
of these papers, the "view" is not a dead bunch of bits (a picture), but an
interactive component/system in its own right.
- Toolglass and Magic Lenses
- Comparing The Programming Demans Of Single-User and Multi-User Applications
- A History-Based Macro By Example System (its deja vu all over again...)
- Bonus paper:Edwards and Mynatt CHI'97 paper on lenses.
For this discussion of Agents, let's focus our discussion on three
articles from last December's IEEE Expert: Intelligent Systems & Their
Applications Special Issue on Intelligent Agents. These articles were
chosen for the following reason: The first paper captures the hype and demand
for agents. The second is a reasonable critique of the approach. The third
offers a number of definitions and techniques. Do these papers all agree
about what an agent is? Is 'agent' well-defined technically, or just
another Holy Grail?
- Papers/Pointers (all from IEEE Expert: Intelligent Systems & Their Applications Special Issue on
Intelligent Agents, Volume 1, Number 6. Published by the IEEE Computer
- Sara Hedberg. Agents for Sale: First Wave of Intelligent Agents Go
Commercial, pp. 16--19.
- Charles Petrie. Agent-based Engineering, the Web, and Intelligence, pp.
- Katia Sycara, Keith Decker, Anandeep Pannu, Mike Williamson, and Dajun
Zeng. Distributed Intelligent Agents, pp. 36--45.
- Notes on discussion
Alternative interfaces to the home
The topic for this meeting is intelligent building design issues, with emphasis
on the HCI issues of interacting with a smart building. (This may not be
apparent from the readings.) We'll discuss many issues, such
as voice/telephony interfaces, but the main points are the problem of
minimizing surface area between user and computer - a.k.a less HCI is
better HCI - and the use of natural affordances in HCI design, or designing
into the environment rather than changing the env to suit the design.
- The Neural Network House
Michael Mozer et al. of the University of Colorado at Boulder are developing
a house which uses a neural network strategy. Their current focus is on
energy conservation and so their system is presently very unidirectional,
computers taking cues from users but not vice versa. Several techniques
used by Mozer's team lend themselves to self-learning systems and to
widely-distributed systems - both key points I will draw on.
- Scott Elrod paper on responsive environments in
Communications of the ACM, July 1993.
Scott touches on privacy and security issues, but a major point
here is the importance of passivity in design - i.e. the user shouldn't
have to explicitly supply information that the computer can get for itself
through remote sensing and the like. Elrod also cites a paper by Loveday,
"Artificial Intelligence for Buildings", published in Applied Energy in
1992 - but this deals almost exclusively with energy management, and has
little relevance to our discussions this quarter.
- Autonomous computing
NASA/Ames has done a great deal of autonomous computing work which has
applicability to intelligent buildings. They introduce the concept of an
"immobot", or immobile robot - i.e. one with many sensors and the ability
to control its "private" world, but with little ability to move around or
interact with the outside world. (Deep space probes, while among the
fastest moving objects in the solar system, are essentially predestined and
watch the universe go by with little control over their path. Their
manipulators are limited to their own extent. Buildings are much the same
in the latter regard.)
- This week's discussion leader invites you to explore his own Web site for an overview of his
house, as well as other hotlinks of interest. After the day's
seminar, we will venture out to see Jonathan's house in person (food included).
- Meeting notes
We've been sidestepping this issue too much. Does ubiquitous computing
mean ubiquitous intrusion?
The paper for this monday is "Architectural Considerations for
Scalable, Secure, Mobile Computing with Location Information" by Mike
Spreitzer and Marvin Theimer. Copies are available outside Gregory's
I'm also going to put a few copies of a blurb that Richard Stallman
wrote in this month's CACM, "The Right to Read". It doesn't entirely
pertain to privacy, but is short and kinda interesting anyway.
Also, if you want, you can check out
- Mike Spreitzer and Marvin Theimer. Architectural Considerations for
Scalable, Secure, Mobile Computing with Location Information. 1994 IEEE International Conference on Distributed Computing Systems, Posman, Poland, 21-24 June 1994, pp. 29-38.
A discussion of how to build in privacy concerns from the beginning.
- Richard Stallman. The Right to Read. Communications of the ACM. 40(2):??, February 1997.
When copyright goes awry.
- Ann CavoukianGo Beyond Security -- Build in Privacy: One Does not Equal the OtherPresented at the ARDTECH/SECURTECH'96 CONFERENCE, Atlanta, Georgia, May 14-16, 1996.
This week, we will talk about videoconferencing, esp. in the context
of design and collaboration. As most of you know, the College will be
splitting into three locations come end of Spring quarter (actually,
we've already split into two locations). Hopefully, these papers will
give us a little insight into the usefulness of having video walls in
each of the three locations and guide us in how to best use them.
Non-traditional 2D interfaces - Lenses
- George Robertson and Jock D. Mackinlay. The Document Lens. In the Proceedings of UIST'93, November, 1993, pages 101-108.
- Scott Hudson and Roy Rodenstein. Debugging Lenses: A New Class of
Transparent Tools for User Interface Debugging. Draft to be submitted
to UIST'97, March, 1997.
Readings for each week will either be included on this page as
links or will be made available a week in advance and announced to the
FCE mailing list.
To subscribe to the mailing list for FCE, send a request to Majordomo@cc.gatech.edu with the line 'subscribe future your-address
(Your Name)' in the body of the message.
For questions regarding this web site or the FCE group, please contact: firstname.lastname@example.org