Proceedings of the Second European Conference on Computer-Supported Cooperative Work
`Bannon, L., Robinson, M. & Schmidt, K. (Editors)
`September 25-27, 1991, Amsterdam, The Netherlands
`
`Idea Management
`In a Shared Drawing Tool
`Iva M. Lu and Marilyn M. Mantei *
`Department of Computer Science
`University of Toronto
`Canada
`* also with the Faculty of Library and Information Science
`
`Abstract
`
`The generation of design ideas in group discussion is a complex and dynamic process.
`Some design ideas are accepted; others are rejected; many others are modified and
`combined. The fluent expression of ideas and the ability to interact and build on
`representations created by others contributes significantly to the idea generation process.
`Computerized shared drawing tools support this fluency and interaction, but such tools
`need to aid not only the drawing process but also the management of design ideas during
`group interaction. This paper lays the groundwork for the design of the idea management
`portion of a shared drawing tool.
`It presents a taxonomy of group idea management
`activities, identifies user requirements in support of these behaviours, and illustrates how
`the user requirements are satisfied by features in CaveDraw, an experimental shared
`drawing system.
`
`1.
`
`Introduction
`
`Because modern technology is complex. it is unusual for an individual to tackle
`the design of a major project single-handedly. Often. a small team is gathered at
`the initial stage of the design process introducing problems of organization.
`coordination and communication. Sketches are an important coordination tool for
`the shared design process and group communication. This group communication
`can be faciliated by computerized shared drawing tools which permit simultaneous
`sketching by team members in different locations. Although these shared drawing
`tools are exceedingly useful. we believe that the management of multiple inputs
`remains a significant issue in their design.
`
`ECSCW'91
`
`97
`
`Petitioners Ex. 1004 Page 1
`
`

`
`Observational studies have identified several critical factors in the design of
`shared drawing tools. These factors are derived from analyzing and interpreting
`collaborative workspace activities. Tang and Leifer (1988) point out that different
`workspace activities occur with different work mediums (e.g., whiteboard, private
`notebooks), different tasks (e.g., mechanics, architecture), and different time-scale
`problems (e.g., multi-year versus two-week projects).
`We believe that understanding the group process of creating and manipulating
`task artifacts -
`sketches in a shared workspace - will allow us to identify user
`requirements in a shared drawing tool. We focus solely on group behaviours as
`members manage and manipulate design ideas and ignore variables like
`cohesiveness and prior design training. Akin (1979) shows that
`the more
`imaginative design alternatives and major design conflicts are often recognized
`while staring at sketches. We believe that supporting group behaviour in
`manipulating the sketches plays a central role in fostering this creativity. We note
`that Grodin (1989) has pointed out that lack of understanding of group behaviour
`is one of the reasons for groupware failure.
`In this paper we are concerned with the design of tools that support idea
`management. Although no direct evidence exists to demonstrate that idea
`management is an important consideration in the design of shared drawing tools,
`we suspect this is an important issue based on empirical evidence from studies of
`individual designers using design aids. Ullman, Stauffer and Dietterich (1987)
`noted that in an individual design session, designers tended to forget some of the
`ideas they formulated. Yeomans (1982) discovered similar recall failures. Ullman
`et al. (1987) also found that a team of designers often worked at different levels of
`abstraction in their design, making it difficult to integrate the final products.
`We also examine studies of group design that did not have the use of shared
`drawing tools. Rouse and Boff (1987) note the following group design behaviour:
`If an outside observer were to characterize designers' behaviors, particularly for complex
`domains such as aircraft design, it is quite likely that such an observer would conclude that
`chaos is the most appropriate characterization of design teams at work.
`They explain the chaos as arising from different design philosophies that
`designers bring to a design team. Scheidel and Crowell (1964) describe group
`decision making as an idea-in-the-making process wherein one member suggests
`an idea, another modifies it and a third changes its focus until the final agreed
`upon solution unfolds. This process of cooperative work in the building of a .group
`decision becomes too complex as more participants are involved. None of the
`above studies indicate that the outcomes of a design are affected by lack of idea
`management and no studies have been done on its use in shared drawing tools.
`However, throughout Section 3, we provide evidence from the literature that
`strongly suggests that the idea management criteria we propose is valid.
`Design ideas are much more than sketches. They also embody task context,
`conversational exchanges, gestures and the order in which all of these take place.
`When we use the term "design idea" we loosely refer to the sketches actually laid
`out on the drawing surface. Thus, we focus on the tasks of choosing, comparing,
`and integrating multiple design sketches. We use these tasks to classify those areas
`
`98
`
`ECSCW'91
`
`Petitioners Ex. 1004 Page 2
`
`

`
`that would benefit from design idea management tools. We propose a set of user
`requirements for the design of multi-user shared drawing tools and illustrate the
`requirements in the design of a prototype, CaveDraw. CaveDraw is a shared tool
`running within a multi-media environment at the University of Toronto (Mantei,
`Baecker, Sellen, Buxton, Milligan & Wellman, 1991).
`
`2. The Approach
`
`To develop our user requirements, we studied videotapes of drawing space
`activities collected by various researchers. We have also drawn on prior research
`in engineering design studies, group communication and social psychology. We
`focused primarily on the interactions between collaborators as they manipulate
`current and previous design ideas. Our research plan is illustrated in Figure 1.
`
`Figure 1. Research focus of this paper
`
`Studying and understanding the scenarios of group behaviour in managing
`design ideas provides us with constraints on how a tool should be designed to
`support them. The scenarios presented in the taxonomy lead us to new insights
`into both shared drawing activity and user requirements for shared drawing tools.
`
`3. A Taxonomy of Group Idea Management Processes
`
`We present our analysis of the design study videos and previous research in the
`form of an idea management taxonomy. The taxonomy is primarily a listing of the
`more general levels of group interchanges and idea manipulation decisions made
`by group members. It is not exhaustive but covers the major behaviours we and
`others have observed in design activity.
`Agree and add on to the suggested idea: A design idea is suggested. One
`or more collaborators make comments on the design either verbally or by
`sketching out the alternatives. Additional sketches are performed to further
`enhance the idea.
`Tang (1989) observed this scenario in his studies; for example, one designer
`(S3) draws a representation of her design idea into the workspace, the other
`designer (SI) builds on the idea by adding keyhole slots. Tang (1989) points out
`
`ECSCW'91
`
`99
`
`Petitioners Ex. 1004 Page 3
`
`

`
`that this behaviour indicates that initial representations gradually evolve into
`distinct artifacts, often through modifications and additions made by others.
`Agree and subdivide the suggested idea: A design idea is suggested.
`Participants agree on the idea as a start. They then proceed to break down
`the idea into sub-tasks or segments and work on them separately.
`Breaking up a design idea into hierarchical sub-tasks is a general phenomenon
`seen in architectural and mechanical engineering design tasks. For instance, in the
`Office Design Project (Stults, 1988), the architects articulated a shared analysis of
`the client's needs, formed a concept (an overall design idea) in response to the
`needs, and summarized the issues (the sub-tasks) underlying the concept.
`Effective management requires that inter-relationships among solutions for each
`sub-task be laid out and saved by the group before the group commences work on
`the sub-tasks (Otto, Riley & Erdman, 1988).
`Modify the suggested idea: A design idea is suggested. One or more
`participants modifies the idea by editing the sketches of the idea or by
`presenting additional related sketches. Participants may not be notified by
`others before their sketch is changed.
`This scenario occurs in studies using Commune, a three-person shared drawing
`tool (Minneman & Bly, 1991). One of the participants erases one of the other
`participant's sketches without requesting prior permission for this action. We
`observed this in a. private viewing of a Xerox PARC design session recorded on
`videotape. Such behaviour is also observed by Tang (1989). He points out that
`the change usually addresses a verbal criticism and such criticism often
`compromises the design idea. In studies of idea development in a small group
`meeting, Scheidel and Crowell (1964) describe how one idea is progressively
`remodified in group interaction until the group achieves agreement.
`Modify, but preserve the suggested idea: A design idea is suggested, and
`participants suggest modifications that are distinct from the original idea.
`These changes can be removed if they don't appear to work.
`Although we did not find this behaviour mentioned in the literature, we
`extrapolate its occurrence from our studies on shared writing (Posner, Baecker &
`Mantei, 1991). Both ForComment™ (Opper, 1988) and Word 4.0™ (Microsoft
`Corporation, 1989) permit this type of annotation in a document without the
`annotation affecting the original text. In the Office Design Project (Stults, 1988),
`one of the architects is observed to lay tracing paper on top of his tv monitor.
`Using another architect's sketch displayed on this monitor, he then proceeds to
`add his own idea on the tracing paper. The original sketch is preserved while the
`other architects comment on the new suggestion.
`Scratch and restart: A design idea is suggested. One or more participants
`comment on the idea, and the originator admits that there is a problem with
`the design idea. The idea is discarded and the group searches for another
`design solution.
`Tang (1989) calls this scenario "Admit Problem". He describes it as one of the
`negotiating patterns in encouraging the group to accept an idea. He notes that this
`
`100
`
`ECSCW'91
`
`Petitioners Ex. 1004 Page 4
`
`

`
`event often encourages others to help resolve the problem and share in developing
`the idea, but that some groups also use the admitted problem to reject the idea. In
`fact, the more ideas that members contribute, the more ideas the group will reject
`(Fisher, 1974). Fisher also points out that the period of idea testing during the
`conflict phase, involves the rejection of many idea proposals.
`Suspend and wait: A design idea is suggested, and one or more
`participants make comments on the idea. Because the group is unsure about
`the suggested idea or because the idea is rejected out of hand, the discussion
`about
`it
`is dropped. The suggested idea can be forgotten or later
`reconsidered in an unrelated context (Tang, 1989).
`In Fisher's (1970) study of decision modification processes in small groups, he
`observes group members introducing a particular decision proposal, discussing it
`for a length of time, dropping it in favour of another decision proposal and then,
`re-introducing it later during the group deliberations.
`Agree and wait: A design idea is suggested and is well received. The
`group moves on to the next sub-task on the requirement list to complete the
`design. The suggested idea is put on hold until all design solutions for the
`overall design are gathered.
`Once a global design idea is agreed upon by participants, it is further broken
`down into design sub-tasks, as mentioned in "Agree and subdivide the suggested
`idea". This scenario is shown in the MacViz-A design studies (Tang, 1989) when
`the participants listed their ideas, one after the other, on the shared workspace. The
`accepted idea was noted and the group moved on to solving the next design issue.
`Compare and consolidate: Multiple design ideas for fulfilling the design
`requirement are suggested. The group compares and criticizes the solutions,
`and then consolidates them into one accepted version. In the consolidation
`process, several design solutions are aborted or modified at the same time.
`Fisher (1974) notes from his studies that
`Group members usually focus their attention on various proposals during their interactions
`and choose from among those alternative proposals the ones which they will accept or
`reject The sum of the proposals accepted constitutes the productivity of the group.
`This type of activity has been observed to occur iteratively whenever a new design
`alternative arises during the design process.
`Deprivatize design idea: After a design idea is generated, it is sometimes
`transferred from an individual workspace to a shared workspace.
`In studies conducted by Tang and Leifer (1988), one participant was observed
`to begin drawing privately, producing a graphical object. Other participants
`noticed the object and began working on it. Tang and Leifer (1988) point out that
`the migration of this object from a private to a public object illustrates the dual
`public/private nature of the workspace.
`We have identified nine distinct design sharing and modification processes that
`have been observed in group design. Naturally, the design process has additional
`complexities and subleties that we have failed to capture. Nevertheless, we believe
`that we have identified some of the primary behaviour patterns that groups apply
`
`ECSCW'91
`
`101
`
`Petitioners Ex. 1004 Page 5
`
`

`
`in managing their design ideas. We need to incorporate capabilities to support
`these patterns in shared drawing tools so that they become facile and fluent
`enough to support this process of developing ideas.
`
`4. A Brief Overview of CaveDraw
`
`taxonomy to generate user
`Before we use the group idea management
`requirements, we provide a brief overview of the shared drawing tool under
`development, CaveDraw. We describe the tool at this point in the paper because
`we will use examples from CaveDraw to demonstrate the application of the user
`requirements we have generat~d.
`CaveDraw is a shared drawing package running on Macintosh II workstations.
`It supports multiple users drawing at the same time. Users working on their
`workstations, connected through an ethernet, can view and modify shared
`drawings in their window. Each workstation runs its own version of CaveDraw
`and communicates with other workstations via a communications manager
`running on a Sun 3/60 workstation.
`CaveDraw differs from other shared drawing software in its support of
`"transparent layers." A layer is created when a user requests and names a drawing
`surface. All users can sketch on the layer. Once a workspace is exhausted on the
`layer, a new layer can be requested. The work on the previous layer dims to a light
`colour so as not to interfere with the drawing on the new layer. Each participant
`can create, hide and select any layer. As layers are superimposed on each other,
`participants can select their own individual layers to work on while drawing
`activity continues by other participants on other layers. Participants can copy or
`cut any portion of a sketch on one layer to a desired location on another layer.
`Sharing a common view of the sketch activity is not automatic in CaveDraw as it
`is in other shared drawing tools. However, participants in CaveDraw can
`synchronize their views with another participant.
`CaveDraw supports line, rectangle, oval, polygon, text, and freehand (pencil &
`marker) drawing tools. Users can select and erase drawing segments and can use
`different coloured markers to identify their own work. CaveDraw shares sketching
`activities but supports gesture weakly. A coloured telepointer is used for gesturing
`but its two-dimensionality will never capture the richness of human gestures.
`CaveDraw is now implemented and is undergoing user testing.
`
`5. User Requirements Drawn From The Group Idea
`Management Taxonomy
`
`In Section 3, we summarized nine idea manipulation behaviours observed in
`groups working on design solutions. We now propose user requirements for the
`design of a shared drawing tool. This tool helps designers manage their design
`ideas. To build these requirements, we combine the nine idea management
`behaviours with five critical factors that have been shown to affect group design.
`As before, our evidence for the importance of these factors comes from prior
`research on shared drawing environments and from reviews of videotapes of
`
`102
`
`ECSCW'91
`
`Petitioners Ex. 1004 Page 6
`
`

`
`shared drawing activities. Once we have discussed user requirements, we present
`CaveDraw's solutions to these requirements. We also review the solutions for
`these requirements in other shared drawing systems and discuss their relative
`merits.
`We focus on five critical factors that support group idea management processes.
`They are Work Allocation, Design Integration, Design Ownership, Design Recall
`and Space Sharing. Work Allocation refers to the split between individual and
`group work. In group design, participants often work on different parts of the same
`design. They therefore need a personal design space that can later be Integrated
`into the group's workspace. Conflict can arise in group design sessions if one
`person's idea is co-opted or erased by another. Thus, Ownership becomes an
`important issue. Drawing space evaporates rapidly as ideas are sketched and
`discarded. Yet, it is important not to eradicate an idea which could be useful in
`another context. If Recall is hindered by the organisation of a design space, prior
`ideas can be lost. Finally, group Sharing of Workspace can limit the amount of
`space available and thus, the number of design ideas generated.
`
`5.1 User Requirements Supporting Work Allocation
`Table I lists the user requirements for the work allocation factor. The "Agree and
`subdivide the suggested idea" scenario generates two requirements: Requirement
`(1.1); to allow participants to select individual segments of the design to work on
`simultaneously, and Requirement (1.2); to provide mechanisms by which each
`participant can be kept aware of what the other participants are doing.
`User ReQUIrements
`axonomv cenanos
`T
`S
`(1.1) Participants can select individual segments of the design
`"Agree and subdivide the
`to work on simultaneously.
`suggested idea"
`(1.2) Participants are aware of the design activities of others
`while workim! on their sel!ment of the desim.
`(1.3) Participants are able to select and modify all previous
`design ideas.
`
`"Modify the suggested
`idea"
`
`Table I. User Requirements Supporting Work Allocation
`We believe that offering participants the choice to work on individual design
`segments simultaneously not only expands the design space for them, but also
`enhances creativity and reduces the processing time of the design task. Thus, they
`can select segments that are relevant to their expertise and work with lower
`communication overhead. Existing instantiations of shared workspaces do not
`permit group participants to retreat and work on a portion of the drawing without
`changing the workspace for the rest of the design group. In Commune (Minneman
`& Bly, 1991), a selection by one participant to move to a previous page of design
`work causes the screens of all participants to be changed to the previous page.
`In a design task, the inter-relationship among design ideas can affect the
`outcome of the overall design. Ullman, Stauffer and Dietterich (1987) observe that
`different designers'
`ideas are sometimes developed at different
`levels of
`abstraction. If members in the group were constantly aware of each other's design
`processes, negotiation and adjustment to an agreed upon standard level of
`
`ECSCW'91
`
`103
`
`Petitioners Ex. 1004 Page 7
`
`

`
`abstraction could go on continuously. Participants could still work on their
`personal design but would be more likely to make it fit into the greater whole.
`Requirement (1.2) therefore requests this awareness capability.
`The "Modify the suggested idea" scenario generates a third user requirement
`(1.3); to permit participants to select and modify all previous designs on an
`individual basis. Manipulating a suggested design idea plays an important role in
`the design process. In the engineering design world, designers attempt one
`solution, move on to a second, then a third, etc. With multiple participants, a large
`number of solution paths are created (Pahl & Beitz, 1984). The group is likely to
`skip a thorough investigation of prior solutions in the interest of group efficiency,
`but the ability to access this work on an individual basis can bring up good ideas
`that would otherwise have been discarded.
`We use shared transparent layers in CaveDraw to implement the user work
`allocation requirements into the design. Their specific relationship to the design
`requirements is shown in Table n.
`UR
`ser (eQurrements
`(1.1) Participants can select individual
`segments of the design to work on
`simultaneously.
`
`CD 'D 'S I '
`ave raws
`esllm o utions
`Participants draw out their design ideas on
`shared transparent layers. Drawings on the
`selected topmost layers are the only ones that
`appear in a brighter colour.
`The layers are superimposed on each other so
`that a participant can see other drawing activities
`taking place in a light grey colour.Participants
`can synchronize their view with the other
`participants. also they can find out who is
`viewing or working on each layer.
`(1.3) Participants are able to select and
`Participants can select, create and hide the
`modify all previous design ideas.
`display of any layer on their screen.
`Table n. CaveDraw Design Features Supporting Work Allocation
`Each CaveDraw participant can create one or more shared layers. The layers
`are stacked together and design sketches on the current working layer are
`displayed in a prominent colour. All the underlying layers are dimmed to a light
`grey colour. Sketches drawn by others are visible but not intrusive. Figure 2
`presents an example of the overlapping layered approach in CaveDraw. In Figure
`2a, Designer A is sketching out her idea of a fIoorplan. The work of Designer B is
`visible but not prominent on the layer below her layer. Figure 2b shows her co(cid:173)
`worker's screen with his layer on top of her layer. He is adding to her work, but
`on a different layer.
`
`(1.2) Participants are aware of the design
`activities of others while working on their
`segment of the design.
`
`104
`
`ECSCW'91
`
`Petitioners Ex. 1004 Page 8
`
`

`
`KoulI!-Dulslllle PI.... 2
`
`Figure 2b. Designer B chooses another
`Figure 2a. Designer A chooses to work
`layer to add to Designer A's idea
`on the fIrst layer where she put her idea
`When participants work separately on different subset of layers, awareness
`of others' work becomes an essential part of coordinating the collaboration. In
`CaveDraw, participants can join or view any other participants' work through the
`View menu. This menu allows participants to view all the selected layers, the top
`layer or any of the dimmed layers of another participant. Also, participants are
`able to restore their own view after browsing through another participant's layers
`or working with another participant. Furthermore, they can find out who is
`working on each particular layer when they select a layer through the Show menu.
`Each menu item gives the name of the layer as well as the name of the other
`participants who are viewing it.
`5.2 User Requirements Supporting Design Integration
`Table III lists the user requirements that support the critical factor, Design
`Integration. Requirement (2.1); to allow participants to compare and consolidate
`modifications to different portions of the original design, while still being able to
`throwaway undesirable changes,
`is a direct result of the "Compare and
`It is not possible to compare complex design alternatives
`consolidate" activity.
`unless they are equally visible. Group design sessions often involve large amounts
`of white paper pinned to walls or the use of a large whiteboard for this purpose.
`Space limitations and the immobility of drawn designs prevent easy comparison of
`distant designs. Commune requires paging through previous designs and bringing
`them up one at a time. VideoWhiteboard (Tang & Minneman, 1991) provides a
`whiteboard sized shared videospace allowing multiple designs to be viewed at the
`same time, but designs are still immovable. In Ishii's (1990) Teamworkstation,
`designs can be overlaid and thus, compared, but the technology limits the number
`of overlays that can be compared in this fashion. Boardnoter in Colab (Stefik,
`Foster, Bobrow, Kahn, Lanning & Suchman, 1987) supports the reduction of
`design alternatives into miniature stamp sheets. The stampsheets can be expanded
`into a full view, but screen space soon exhausts the number of expanded
`stampsheets that can be viewed at one time.
`Consolidating designs is even less easy. Separate designs have to be redrawn
`and re-merged into a new design requiring a duplication of effort.
`
`ECSCW'91
`
`105
`
`Petitioners Ex. 1004 Page 9
`
`

`
`Teamworkstation allows participants to overlay video images of separate drawings
`but the adjustments require fine tuning using video controls. The consolidated
`design becomes the video sequence stored on videotape.
`User ReaUIrements
`Taxonomy Scenarios
`(2.1) Participants can,with little overhead, compare and consolidate
`"Compare and
`modifications to different portions of the original design but still
`consolidate"
`throw out undesirable changes.
`(2.2) Participants can compare different modifications to a design
`idea at the same time without disturbing the original idea or having to
`view multiple disnlavs.
`(2.3) Participants can, with little effort, view both the overall design
`and its subunits in addition to the design subunit they are working on.
`
`"Modify but preserve
`the suggested idea"
`
`"Agree and subdivide
`the suggested idea"
`
`Table III. User Requirements Supporting Design Integration
`The "Modify, but preserve the suggested idea" scenario creates Requirement
`(2.2); to allow participants to compare modifications to a design idea without
`disturbing the original. In a group design session, participants may have an agreed
`upon basis for their design, but may be trying out additional ideas to correct some
`aspect of the design. For example,
`they may want to design the lighting
`connections in the trunk of a new car, while retaining the trunk cavity layout
`design. If they drew over the trunk cavity design on the whiteboard and did not
`like the design idea, they would need to redraw the trunk cavity.
`Requirement (2.3); to allow participants to view both the overall design and all
`its subunits, is drawn from the "Agree and subdivide the suggested idea" scenario.
`As participants create their solutions, they move further away from their original
`plan. Suchman and Trigg (1986) point out that participants relate their ideas to
`prior ones or to the problem at hand. If the original plan is not viewable from time
`to time, participants relate their current problem to the most recently solved
`problem. This eventually places designs far enough away from the overall design
`that integration could be very difficult. If multiple designers work individually
`without refering to the overall plan, their designs are unlikely to fit together.
`The CaveDraw layer approach allows a form of design integration although it,
`too, has limitations. Table IV lists the CaveDraw design features that support the
`user requirements of Design Integration.
`eQUIrements
`U R
`ser
`(2.1) Participants can,with little overhead,
`compare and consolidate modifications to
`different portions of the original design but still
`throw out undesirable changes
`
`C D 'De" Sl'
`ave raws
`sum o utions
`Allows the participants to draw alternate design
`ideas on different layers and superimpose the
`layers or subsets of the layers in any order
`selected by the participants. Also allows saving
`of any of these combinations.
`(2.2) Participants can compare different
`Same approach as (2.1). In addition,
`modifications to a design idea at the same time participants can work on their own layer while
`without disturbing the original idea or having
`the other participants are performing a
`to view multiple displays.
`comparison.
`(2.3) Participants can, with little effort, view
`Allows each participant to bring up a sublayer
`both the overall design and its subunits in
`showing the connection of all subunits while
`addition to the design subunit they are working working on one of the subunits in the previous
`on.
`laver.
`Table IV. CaveDraw Features Supporting Design Integration
`
`106
`
`ECSCW'91
`
`Petitioners Ex. 1004 Page 10
`
`

`
`To support comparison and consolidation activItieS, CaveDraw permits
`participants to select a subset of layers to be displayed on their screen. The designs
`in this subset can be compared and, if desired, consolidated into one design on one
`layer. Design ideas can be discarded by removing the layer on which they are
`drawn. Any layer can be brought to the topmost position (brighter colour) by
`"mouse-clicking" on a pixel located in the layer. Layers below the topmost layer
`are still visible for the comparison task.
`If an overview of the design plan is available, participants can maintain the
`overview as one of the visible layers on their display. They can refer to it from
`time to time by bringing it to the topmost layer or simply by looking at it through
`the other designs showing in the other layers.
`Although CaveDraw supports some of the characteristics of Design Integration,
`it leads to what we call "layer overload." At some point, too many designs with
`too many different patterns will overlap each other in the layered design space. It
`will be difficult for users to disambiguate the lines of one layer from that of the
`other. A feature paralleling Furnas's (1986) fisheye views approach of looking at a
`design overview would be more useful for this function.
`
`5.3 User Requirements Supporting Design Ownership
`Ideas have creators and thus, owners. Any time a sketch is modified by other
`participants in the group, ownership preservation becomes an issue. The design
`scenarios, "Add on to," "Modify," and "Deprivatize," represent different ways in
`which an existing idea can be co-opted by the group. Table V lists user
`requirements that preserve ownership: Requirement (3.1); to allow participants to
`declare any portion of a sketch as private and therefore, undeletable, and
`Requirement (3.2); to allow participants to see who is working on what design.
`Taxonomv Scenarios
`User ReQuirements
`"Agree and add on to the
`(3.1) Participants can declare any portion of a sketch as
`suggested idea", "Modify the
`private and not subject to deletion by others.
`suggested idea", and "Deprivatize
`(3.2) Participants can identify, with no additional
`interaction sequences, who is working on any specific
`design idea"
`design sketch.
`
`Table V. User Requirements Supporting Design Ownership
`Social norms are expected to keep others from erasing our work, but this does
`not always work. For instance, one dominant participant using Commune was
`observed to erase the other person's sketches without prior permission. Conflict
`resolution studies using the University of Minnesota's Group Decision Support
`System found that asocial acts of removing another participant's ideas were
`common and disturbed the group process (Poole, Holmes & DeSanctis, 1988).
`Ownership prevents undesired deletion of design ideas, but sometimes deletion
`or permission to copy is desired. If Requirement 3.1 is met, then Requirement 3.2
`needs to be in place to identify the owner of the design. Identified owners can then
`be asked if deletion or duplication is acceptable. Individuals in a design group may
`also have status. For example, it may not be obvious to other participants that a
`particularly complex design idea is a good solution, but if it is known that the
`
`ECSCW'91
`
`107
`
`Petitioners Ex. 1004 Page 11
`
`

`
`person who created the idea has a reputation as an extremely successful designer,
`then evaluation of the design will be more positive. It is also important to know
`who designed what in a design p