The second part of this paper starts with the history of computer workspaces. In fact, I've never seen smalltalk projects or NLS. But it was easy to understand what multiple shared desktops are and why they are needed. (Because I use similar window manager on Sun workstations everyday.) The authors say that 'Information Visualizer' is more than just multiple shared desktop environments. They say that their goal is to evolve the Rooms multiple desktop metaphor into a 'information workspace' that treats the complete cost structure and considers information access part of a larger work. I think this idea is fairy good but somewhat too abstract. I can realize that there is difference between multiple shared desktop and information workspace. But this paper does not include enough explanation about which feature of 'Information Visualizer' makes Room metaphor into 'information workspace'.
But for major three components of 'Information Visualizer', this paper contains detail explanation. The main idea of the system is summarized as Table 2, which I liked best in this paper. This table represents what the reader should understand in a clear format. General theories and corresponding system goals are listed by observations. Then the authors identify which part of the system solves each problem. The following part contains the detail rationale of the design of 'Information Visualizer'.
In discussion part, after a brief summary of the system, the writer talks about systems research paradigm. They say that they have been able to utilize theory and empirical relationships of previous researches as well as general theoretical observations from 'tools for thought'.
This paper shows how to combine a new user interface and the analytical and empirical foundation. I think this paper also can be a good sample how UI designer evaluates and justifies his design.
Each piece of information has a cost associated with finding and
accessing it.
Immediate storage, Secondary storage, Tertiary storage
Six Observations derived from studies of Information processing systems
Hierarchy:
High Cost Ratios: from less than a second to more than an hour
Locality of Reference: working set
Reference Clustering: abrupt transition among working sets
Max Info/Cost:
Abstraction: cut down the volume of information to be processed by the user
through more abstract and simpler representations.
Providing some sort of low-cost, immediate storage for accessing objects
in use. A special environment in which the cost structure is tuned to the
requirement of the work processes using them.
Evolution of workspace
Point & Click Editor NLS
Desktop Metaphor Smalltalk
Large Desktop Big Screen Dataland
Multiple Desktops Smalltalk projects
Multiple Shared Desktops Rooms
Information Workspace Information Visualizer
Information Workspace
; To evolve the Rooms metaphor into a workspace that
1) treats the complete cost structure
2) consider information access part of a larger work processes
Information Visualizer
; experimental system
1) 3D/Rooms
Locality of reference: local workspace - room
Reference clustering: multiple rooms
Make the Immediate storage larger by having multiple desktops
Make the Immediate storage denser - Active process (by J.J. Gibson)
The user seems to be able to be aware of objects that are occluded
if he can easily (within about a second) update his knowledge of them
Walking metaphor: animation-based UI
2) Cognitive Co-process
; rapid cycles of the user-system interaction
UI interaction manager derived from Sheridan's notion of supervisory
control - The user is trying to control multiple applications running
as semi-autonomous agents.
Perceptual processing: the screen must be repainted at least once
every 0.1 sec
Immediate processing: status feedback at intervals no longer than
some constant
Unit task: the user can begin the next request as soon as sufficient
information has developed.
3) Information visualizations
Display structural relationships and context
Hierarchical structure
Linear structure - time
Scientific data (see Card Plate 1 - Card plate 2 in the paper is a typo)
Office Building (see Card Plate 2)
Rooms overview (see Card Plate 3)
Indexing and Searching
TDB (subsystem) provides stemming and a full text inverted database.
Documents = word vectors
Desktop metaphor a 3D/Rooms, Cognitive Co-processor, Information visualization To integrate Theoretical and Empirical analysis with practical system design System research paradigm 1) Initial exploratory design - incrementally improved 2) Abstractions 3) Theories and observations to characterize the design 4) Knowledge of the design space is codified