TeleNotes: Managing Lightweight
`Interactions in the Desktop
`
`STEVE WHITTAKER, JERRY SWANSON, JAKOV KUCAN, and
`CANDY SIDNER
`Lotus Development Corporation
`
`Communication theories and technology have tended to focus on extended, formal meetings
`and have neglected a prevalent and vital form of workplace communication—namely, light-
`weight communication. Unlike formal, extended meetings, lightweight interaction is brief,
`informal, unplanned, and intermittent. We analyze naturalistic data from a study of work-
`place communication and derive five design criteria for lightweight interaction systems. These
`criteria require that systems for lightweight interaction support conversational tracking,
`rapid connection, the ability to leave a message, context management, and shared real-time
`objects. Using these criteria, we evaluate existing interpersonal communications technologies.
`We then describe an implementation of a system (TeleNotes) that is designed to support
`lightweight interaction by meeting these criteria. The interface metaphor allows communica-
`tions to be based around desktop objects, resembling “sticky notes.” These objects are also
`organized into “desktop piles” to support conversational threads and provide mechanisms for
`initiating real-time audio, video, and application sharing. We conducted informal user testing
`of several system prototypes. Based on our findings, outstanding issues concerning theory and
`systems design for communication systems are outlined—in particular, with regard to the
`issue of managing conversations over time.
`Categories and Subject Descriptors: H.1.2 [Model and Principles]: User/Machine Systems—
`human factors; H.5.1 [Information Interfaces and Presentation]: Multimedia Information
`Systems — evaluation/methodology; H.5.2 [Information Interfaces and Presentation]:
`User Interfaces — evaluation/methodology; H.5.3 [Information Interfaces and Presenta-
`tion]: Group and Organizational Interfaces—asynchronous interactions; evaluation methodol-
`ogy; synchronous interactions; theory and models; I.3.6 [Computer Graphics]: Methodology
`and Techniques—interaction techniques
`General Terms: Human Factors
`Additional Key Words and Phrases: Audio, awareness, computer-media spaces, conversation
`management, impromptu communication, informal communication, interpersonal communica-
`tions, lightweight interaction, mediated communication, remote collaboration, task manage-
`ment, video
`
`Authors’ addresses: S. Whittaker, ATT Labs—Research, 180 Park Avenue, P. O. Box 971,
`Florham Park, NJ 09732; email: whittaker@research.att.com; J. Swanson, J. Kucan, and C.
`Sidner, Lotus Development Corporation, One Rogers Street, Cambridge, MA 02142.
`Permission to make digital / hard copy of part or all of this work for personal or classroom use
`is granted without fee provided that the copies are not made or distributed for profit or
`commercial advantage, the copyright notice, the title of the publication, and its date appear,
`and notice is given that copying is by permission of the ACM, Inc. To copy otherwise, to
`republish, to post on servers, or to redistribute to lists, requires prior specific permission
`and / or a fee.
`© 1997 ACM 1073-0516/97/0600 –0137 $03.50
`
`ACM Transactions on Computer-Human Interaction, Vol. 4, No. 2, June 1997, Pages 137–168.
`
`Ebates Ex. 1013, Page 1 of 32
`Ebates Performance Marketing, Inc. v. International Business Machines, Inc.
`IPR2022-00133
`
`

`

`138
`
`•
`
`Steve Whittaker et al.
`
`1. LIGHTWEIGHT COMMUNICATIONS
`interpersonal
`Studies of workplace activity show the pervasiveness of
`communications. Office workers spend between 25% and 70% of their time
`in face-to-face conversations with others, depending on job specification
`[Kraut et al. 1993; Panko 1992; Whittaker et al. 1994]. In addition to being
`frequent, interpersonal communication is also vital for achieving certain
`types of work-related tasks. Frequent, opportunistic face-to-face communi-
`cations are crucial for rapidly resolving ambiguity during the planning and
`negotiation phases of projects, and they support organizational learning
`[Finholt et al. 1990; Kraut and Streeter 1996; Kraut et al. 1990; 1993;
`Suchman and Wynn 1984]. However, apart from a few recent exceptions
`[Bly et al. 1993; Fish et al. 1992; Gaver et al. 1992; Isaacs et al. 1996; Tang
`et al. 1994], past systems and theoretical work have tended to focus on only
`one class of interpersonal communication. This is the class of interactions
`that are extended, multiparty, formal, and one-shot [Egido 1988; 1990;
`Johansen 1984; Nunamaker et al. 1993; Olson et al. 1992; Stefik et al.
`1987; Tang 1991]. Recent research shows, however, that such extended,
`multiparty, formal, one-shot interactions are the exception, rather than the
`rule, in interpersonal communications. The majority of office interactions
`consist of brief informal two-person exchanges, for example, to answer a
`colleague’s question, to remind a coworker about a deadline, to hand over a
`document, or to discuss a social issue [Frohlich and O’Conaill 1995; Isaacs
`et al. 1997; Kraut et al. 1993; Suchman and Wynn 1984; Whittaker et al.
`1994]. Here we therefore focus on brief, two-person, informal, repeated
`communications which we call lightweight interactions.
`Research on scientific collaboration has shown that physical distance is
`the strongest predictor of whether two researchers will collaborate, pre-
`cisely because physical proximity promotes lightweight interaction [Kraut
`et al. 1990; 1993]. The data we report below show exactly how a shared
`physical environment leads to these types of interaction. However, trends
`toward telework, mobile work, and the globalization of business are geo-
`graphically separating workers, making collocated lightweight interaction
`less frequent and harder to achieve. In addition the use of technologies
`such as email and workflow may be decreasing the lightweight interaction
`that in the past took place around face-to-face document distribution and
`meeting scheduling [Whittaker 1996].
`What is needed to support effective lightweight interaction at a distance?
`Technologies to support lightweight interaction will need to support its key
`features. Empirical work shows interpersonal communication1 is (a) usu-
`ally two-person rather than multiparty, so that for professional workers
`
`1The data we report here cover all forms of interaction and not just lightweight interaction.
`They therefore include instances of more extended, formal, arranged meetings. By including
`all forms of interaction, these summary data thus underestimate some of the major character-
`istics of lightweight interaction—namely, their dyadic nature, brevity, opportunism, and
`repeated nature.
`
`ACM Transactions on Computer-Human Interaction, Vol. 4, No. 2, June 1997.
`
`Ebates Ex. 1013, Page 2 of 32
`Ebates Performance Marketing, Inc. v. International Business Machines, Inc.
`IPR2022-00133
`
`

`

`TeleNotes: Managing Lightweight Interactions in the Desktop
`
`•
`
`139
`
`84% of meetings are dyadic [Whittaker et al. 1994]; (b) brief, with conver-
`sations generally lasting no more than a few minutes [Kraut et al. 1993;
`Whittaker et al. 1994];
`(c) opportunistic rather than scheduled, with
`professionals having around 90% unscheduled meetings [Kraut et al. 1993;
`Whittaker et al. 1994], with figures of around 60% unscheduled meetings
`for managers [Panko 1992]; (d) focused around shared objects such as
`documents or designs [Luff et al. 1992; Tang 1991]; in our data we found
`that documents were involved in 53% of interactions [Whittaker et al.
`1994]; and (e) repeated rather than one-shot. Repeated communications are
`often necessary because the purposes of lightweight interactions are seldom
`achieved in one interchange, with the result that such conversations are
`frequent and intermittent [Whittaker et al. 1994], with participants on
`average interacting with each other 2.5 times per day [Kraut et al. 1993].
`To illustrate the character of lightweight interaction, we present two
`illustrative examples, taken from an extensive analysis of real workplace
`interactions. The data come from an observational study in which knowl-
`edge workers were “shadowed” for a week. We recorded all their interac-
`tions using a combination of video and audio. We generated a corpus of 294
`conversations, involving a total of 99 different interactants. This research
`method and more detailed results are presented in Whittaker et al. [1994]
`and Frohlich [1995]. Both the current examples involve the same two
`participants, R and F, who work together in a surveyor’s2 office.
`In Example 1.1, the entire two-person interaction is completed in four
`utterances and lasts only eight seconds. R sees that F is moving around the
`office and hence is not currently engaged directly in work. R therefore
`opportunistically solicits F’s help. The fact that they share a common
`physical environment affords R this information about F’s availability and
`allows F and R to jointly look at, and then physically exchange, the
`document. Note also the brevity of the interaction and the absence of
`formal conversational openings or closings, such as greetings or farewells.
`Example 1.1. A short opportunistic interaction eliciting feedback about
`a document (duration: eight secs.).
`R IS STANDING UP READING A DOCUMENT BEHIND HIS DESK
`WHEN HIS COLLEAGUE F WALKS INTO VIEW. F IS ON HIS WAY
`TO HIS OWN DESK FROM ANOTHER OFFICE.
`1. R: “F, can you read this report for me?”
`2. F: “Now?”
`3. R: “Aye if you’ve got a minute.”
`4. F: “Yeah.”
`Example 1.2. An opportunistic interaction leading to unsolicited advice
`(28-second fragment of conversation lasting 1:36 mins.).
`
`2In U.S. English, a surveyor is referred to as a real estate appraiser.
`
`ACM Transactions on Computer-Human Interaction, Vol. 4, No. 2, June 1997.
`
`Ebates Ex. 1013, Page 3 of 32
`Ebates Performance Marketing, Inc. v. International Business Machines, Inc.
`IPR2022-00133
`
`

`

`140
`
`•
`
`Steve Whittaker et al.
`
`F IS ON THE PHONE ACROSS THE OFFICE FROM R. AS SOON AS
`F PUTS DOWN PHONE, R BEGINS TO SPEAK.
`
`1. R: “Is he alright?”
`2. F: “Yeah”
`3. R: “Which one’s he’s got . . . there’s a restaurant”
`4. F: “I said that I’ll do this one initially and then further afield”
`5. R: “Which one’s that?”
`6. F: “That’s 82 Whiteladies Road; it’s the offices”
`7. R: “Oh, yeah we act for the landlord on that one. I did a rent review
`against him on that”
`8. F: “Right”
`9. R: “His shop it might be worth checking out. He’s got a subtenant
`downstairs who’s got a clothes shop”
`10. F: “Yeah”
`11. R: “Might be worth trying to get in with them as well”
`12. F: “Yeah alright”
`
`The fragment in Example 1.2 represents 28 seconds of a second interac-
`tion lasting 96 seconds. Although it again features R and F, it shows an
`unplanned conversation about a different topic, one that occurred immedi-
`ately after F had finished a phone call to a client. It arose spontaneously
`because R heard F’s phone call and opportunistically wanted to monitor the
`outcome (line 1). Again the fact that participants share a common physical
`environment promotes this type of impromptu interaction. It continues
`with R offering unsolicited advice and assistance. Again the interaction
`starts without formal initiation, with R beginning with a direct question to
`F. The interaction also has a history, as revealed by the implicit shared
`context between the participants: without being told, R knows the identity
`of F’s caller (line 1) and details of the case (line 3), although R cannot
`remember all these details. R also proceeds to supply background informa-
`tion, which F may not already have known (line 7), and some advice (lines 9
`and 11). The shared context results in a condensed and cryptic conversa-
`tional style. This conversation continued for several more utterances after
`the extract: R gave more details and offered a warning about acting for
`both client and tenant. Thus an unplanned conversation led to a detailed
`task-oriented discussion initiated by R that was of benefit to F. Further
`analyses of these conversations, along with a demonstration of their
`functions and mutual benefits to the interactants, are given in Frohlich
`[1995] and O’Conaill and Frohlich [1995].
`These results suggest a characterization of work in which people are
`engaged in multiple, intermittent, interleaved collaborative tasks. Workers
`seek out and are frequently sought out by their coworkers for brief
`synchronous opportunistic interactions. Each conversation may have a
`history of prior interactions, and workers are often concurrently engaged in
`multiple interaction threads. Furthermore, a worker may be engaged in
`
`ACM Transactions on Computer-Human Interaction, Vol. 4, No. 2, June 1997.
`
`Ebates Ex. 1013, Page 4 of 32
`Ebates Performance Marketing, Inc. v. International Business Machines, Inc.
`IPR2022-00133
`
`

`

`TeleNotes: Managing Lightweight Interactions in the Desktop
`
`•
`
`141
`
`multiple concurrent conversational threads with the same coworker. This
`gives rise to the problems of context regeneration and tracking conversa-
`tional threads. Given that workers are engaged in multiple concurrent
`tasks, and that conversations are usually impromptu, how are workers able
`to switch context from their current task and immediately refocus on the
`topic that their coworker has just asked about? Furthermore, when the
`brief interaction is over, how do they switch back to their prior interrupted
`task [O’Conaill and Frohlich 1995]? How also do they keep track of the
`relationships between different fragments of the same intermittent conver-
`sation, when there are often long delays between related interactions?
`We will argue that people currently exploit the presence of work-related
`artifacts (such as papers, drawings, notes, and folders) to help manage the
`history and context of these intermittent interactions. When task-related
`documents are copresent and visible, they can serve to “hold the context” of
`multiple ongoing conversations. Indirect support for this notion comes from
`our observation that 53% of workplace interactions involved a document
`[Whittaker et al. 1994]. Additional evidence for the context management
`function of documents is the frequently observed “messy desktop” [Barreau
`and Nardi 1995; Frohlich 1995; Kidd 1994; Malone 1984; Mander et al.
`1992; Rouncefield et al. 1994]. Office workers scatter their physical desk-
`tops with clusters of notes, documents, and folders relating to ongoing
`projects. As a consequence, this information is readily at hand as a
`retrieval cue when an external interruption occurs. The “messy desktop”
`therefore allows people to regenerate the history of a prior conversation,
`even when substantial time has elapsed since the topic was last discussed.3
`Other work has described people’s attempts to use the computer desktop
`in an analogous way: electronic documents and folders are sometimes left
`visible in the user’s electronic workspace to serve as reminders and
`context-holders for urgent work in progress [Barreau and Nardi 1995].
`Similarly, users often retain undischarged email messages in their inboxes
`to serve as reminders about tasks or conversations that are still in progress
`[Whittaker and Sidner 1996]. Despite this, few systems provide direct
`support for context maintenance.4 Furthermore, a number of recent studies
`have emphasized the utility of paper documents in providing an external
`visible record of current ongoing activity and have contrasted this with the
`relative inaccessibility of screen-based information for context tracking
`[Bowers 1994; Luff et al. 1992; Whittaker and Schwarz 1995].
`
`3An alternative perspective is that the “messy” desktop is actually malfunctional, resulting
`from a lack of time or inclination to carry out filing and that it is an ineffective way to organize
`ongoing tasks. While we know of no study that has quantitatively compared the success of
`“messy” and “neat” desktops, there is strong evidence of the reminding functions of visible
`materials [Whittaker and Schwarz 1995; Whittaker and Sidner 1996], as well as the ability of
`users to find information in “messy” piles [Barreau and Nardi 1995].
`4One exception is a system by Henderson and Card [1986] that uses the metaphor of “rooms”
`into which users can place related documents or applications, where it is easy to switch rooms
`and hence contexts.
`
`ACM Transactions on Computer-Human Interaction, Vol. 4, No. 2, June 1997.
`
`Ebates Ex. 1013, Page 5 of 32
`Ebates Performance Marketing, Inc. v. International Business Machines, Inc.
`IPR2022-00133
`
`

`

`142
`
`•
`
`Steve Whittaker et al.
`
`These paper documents and notes also serve a second role. Once an
`interaction is underway they function as shared objects. They operate as
`part of a shared workspace and provide a shared physical context for the
`conversational participants. They also act as (1) a resource for looking up
`and recording information and (2) as a target for gesturing, marking, and
`note-taking [Frohlich 1995; Luff et al. 1992; Tang 1991; Whittaker et al.
`1994]. Studies of remote synchronous communication using audio and
`shared workspace have also documented these functions for electronic
`documents and shared materials as “context-holders” and memory aids in
`real-time discussions [Bly 1988; Brinck and Gomez 1992; Minneman and
`Bly 1991; Whittaker 1995a; Whittaker et al. 1991; 1993]. Similar benefits
`occur when groups are provided with a shared dynamic image to support
`collaboration at a distance [Gaver et al. 1993; Nardi et al. 1993; 1996;
`Whittaker 1995a; 1995b]. In these applications video supports a shared
`workspace by providing joint access to video images of work objects that are
`critical to the collaborative task of the distributed group. In a neurosurgery
`application [Nardi et al. 1993; 1996; Whittaker 1995a; 1995b], remote
`members of a distributed surgical team were able to view video images
`from the operating theater of the patient’s brain or spine. They were
`therefore able to see the state of the operation “at a glance” by looking at
`the image. This enabled them to coordinate their remote actions and
`provide advice to the team in the operating theater.
`Using artifacts as reminders and context-holders for ongoing tasks is also
`important in the event of failed attempts to initiate lightweight interaction.
`We refer to this as the connection problem. The opportunistic nature of
`lightweight interaction means that attempts to initiate communication
`often fail, because the intended conversational participant is not currently
`available for conversation. Thus we found that attempts to initiate im-
`promptu communications using the telephone failed on 62% of occasions,
`because the intended recipients were away from their desks or otherwise
`engaged in communicating with another person [Whittaker et al. 1994].
`Similar failure rates are reported by other studies [Rice and Shook 1990].
`This suggests two new communication requirements: one-way drop and
`communication reminders. For many communication purposes, one-way
`information transmission may be sufficient: surveys report that users feel
`that leaving a message is adequate for over half of business telephone
`conversations [Rice and Shook 1990]. For this reason, office workers often
`leave each other brief notes accompanying documents (“here’s the most
`recent draft; can you look at pages 3–5?”). On other occasions, however,
`one-way drop of information may be insufficient. It may therefore be
`necessary to have synchronous communication, in which case a communi-
`cation reminder may be important: leaving a message (“call me before 11,
`on 123 4567”) can be used to coordinate a future synchronous connection
`[Tang et al. 1994]. Following from this, a further function of desktop notes,
`documents, and folders may also be to remind the worker of the fact that a
`connection attempt failed and that a synchronous conversation still needs
`to take place. Furthermore, artifacts can hold context during the period
`
`ACM Transactions on Computer-Human Interaction, Vol. 4, No. 2, June 1997.
`
`Ebates Ex. 1013, Page 6 of 32
`Ebates Performance Marketing, Inc. v. International Business Machines, Inc.
`IPR2022-00133
`
`

`

`TeleNotes: Managing Lightweight Interactions in the Desktop
`
`•
`
`143
`
`that one is waiting to communicate synchronously with another person.
`Thus when the connection is successfully made, the original conversational
`context can be more easily regenerated, because the presence of the
`documents serves as a retrieval cue for memory access.
`This analysis of workplace communications indicates that effective sup-
`port for lightweight interaction requires five related components:
`(a) Conversational threading: Participants are engaged in multiple inter-
`mittent communications tasks, often with different individuals. The
`system must therefore keep track of interactions, storing elements of
`the same conversation together so that they can rapidly be accessed as
`a unit, allowing participants to check the status of a given conversation.
`(b) One-way drop: The system should support the ability to leave a brief
`asynchronous message, given the fact that attempts to achieve opportu-
`nistic connections frequently fail, and a valuable-information exchange
`can often take place without synchronous communications.
`(c) Quick connection: Given the brevity of lightweight interaction, the
`system should support rapid flexible communications, and participants
`should be able to quickly connect with others.
`(d) Context preservation and regeneration: Given the intermittent but
`repeated nature of lightweight interaction, where there are often long
`delays between elements of the same conversation, the system should
`support methods for straightforwardly accessing prior parts of ongoing
`conversations, including the materials or artifacts that are involved in
`that interaction.
`(e) Shared objects: The system should support real-time shared objects as
`props and conversational resources, both because of the frequency with
`which documents feature in lightweight interaction and their support-
`ing role in mediating conversation.
`Principles (a) and (d) and to a lesser extent (b) require the management of
`interactions across time, a problem which has characteristically been
`ignored in theoretical, empirical, and technology-oriented work [O’Conaill
`and Frohlich 1995; Whittaker and Sidner 1996; Whittaker et al. 1994]. In
`contrast, (c) and (e) are more concerned with real-time interaction, which
`has received more attention. Overall we need to support both synchronous
`and asynchronous communication methods, as part of an integrated set of
`communications applications for lightweight interaction [Tang et al. 1994].
`Users can then choose the appropriate communication method as the
`situation demands.
`
`2. CURRENT TECHNOLOGIES FOR SUPPORTING LIGHTWEIGHT
`COMMUNICATIONS: STRENGTHS AND WEAKNESSES
`Our discussion has mainly focused on analyzing lightweight communica-
`tions in workgroups who share the same physical location. We now evalu-
`ate current technologies used for remote communication to see how well
`they support the lightweight-interaction characteristics identified above.
`
`ACM Transactions on Computer-Human Interaction, Vol. 4, No. 2, June 1997.
`
`Ebates Ex. 1013, Page 7 of 32
`Ebates Performance Marketing, Inc. v. International Business Machines, Inc.
`IPR2022-00133
`
`

`

`144
`
`•
`
`Steve Whittaker et al.
`
`Table I. Evaluating Current Communications Technologies
`for Lightweight-Interaction Features
`
`Type
`
`Task
`Threading
`
`Tele-
`phone
`
`Voice-
`mail
`
`Email
`
`No
`
`No Weak support
`in some
`systems
`Yes
`
`Yes
`
`One-Way Drop
`
`No
`
`Yes
`
`Yes
`
`Work-
`flow Pager
`
`Media
`Spaces
`
`Yes
`
`No
`
`No
`
`Videophone/
`Videocon-
`ference
`
`Shared
`Work-
`spaces
`
`No
`
`No
`
`No
`
`No
`
`Yes
`
`Yes
`
`Not
`supported
`in most
`systems
`Yes
`
`Yes
`
`No
`
`No
`
`No
`
`No
`
`Quick
`Synchronous
`Connect
`Context
`Regeneration
`
`Real-Time
`Shared
`Objects
`
`No
`
`No
`
`No Weak support
`in some
`systems
`No
`
`No
`
`Yes
`
`No
`
`No
`
`No
`
`No
`
`No
`
`Not
`supported
`in most
`systems
`
`Not
`supported in
`most
`systems
`
`No
`
`Yes
`
`Table I shows that no current technology supports all the features of
`lightweight interaction. Technologies tend to focus on purely synchronous
`communications (e.g., telephone, shared workspace, videophone) without
`support for context regeneration or task threading. Alternatively, they are
`focused on asynchronous communication (e.g., email, workflow, and voice-
`mail), which is highly effective for information drop-off, but does not
`support real-time exchanges. Furthermore, we will argue that with the
`possible exception of workflow, most asynchronous technologies do not
`support context regeneration and task threading effectively.
`The telephone supports synchronous connection, but it has no document
`integration or asynchronous components. It therefore provides no support
`for managing the history of an intermittent extended conversation by
`offering features such as context regeneration and task threading. It also
`provides no capability for real-time object sharing. The telephone alone
`does not support one-way drop of information, although combined with an
`answering machine it does so.
`Voicemail does not support synchronous communications or real-time
`shared objects. Voicemail supports one-way drop of information, often as a
`means to promote synchronous communications, e.g., “this is X; call me
`back on 123 4567.” Although there is message storage in voicemail, most
`systems limit storage to small numbers of messages which have to be
`deleted on a frequent basis,5 and no commercial systems provide users with
`
`5Part of the problem here is that the size of audio files makes it costly to store large numbers
`of messages, which precludes the creation of message archives.
`
`ACM Transactions on Computer-Human Interaction, Vol. 4, No. 2, June 1997.
`
`Ebates Ex. 1013, Page 8 of 32
`Ebates Performance Marketing, Inc. v. International Business Machines, Inc.
`IPR2022-00133
`
`

`

`TeleNotes: Managing Lightweight Interactions in the Desktop
`
`•
`
`145
`
`ways to manage messages according to thread [Rice and Shook 1990].
`Again this means that the prior context of a voicemail message is often lost.
`Email does not support synchronous connection. It also does not allow
`users to share objects in real time. Given its asynchronous nature, email
`supports one-way drop of information. Most email systems also provide
`minimal features for context regeneration, in that they allow users to save
`messages. However, the majority of systems lack features that explicitly
`facilitate context management, such as the automatic categorization of
`messages by conversational task. For example, one might wish to view
`originating messages and responses to those messages together to deter-
`mine the conversational context for a response one is generating or to track
`the state of an ongoing conversational thread [Whittaker and Sidner 1996].
`Most email systems provide folders, for the categorization of messages, and
`hence provide weak support for context management, but categorization
`requires the user to execute each action manually, rather than being an
`intrinsic system feature.6
`Workflow systems do not allow synchronous connection, or shared objects,
`but their primary aim is to allow users to track the set of interactions that
`are associated with complex transactions occurring over extended periods
`of time. They therefore support aspects of threading and context regenera-
`tion [Abbott and Sarin 1994; Winograd and Flores 1986]. Workflow systems
`are asynchronous and in principle support one-way drop, although they
`tend not to be used this way, as they are intended to manage more
`extended interactions.
`Pagers do not support any of the above features, being solely a means to
`drop information, such as a short message or phone number. They do not
`support any other communication features.
`Media spaces are recently prototyped technologies that provide users
`with permanently open video/audio links or methods to quickly establish
`synchronous video/audio links [Bly et al. 1993; Fish et al. 1992; Gaver et al.
`1992; Mantei et al. 1991; Tang et al. 1994]. Recent systems provide some
`integration with real-time shared applications to allow object sharing7
`[Tang et al. 1994]. Two systems [Gaver et al. 1992; Tang et al. 1994] also
`provide support for one-way drop, but none of these systems helps with
`context regeneration or task threads.
`Videophones (and their multiparty equivalent, videoconferencing) are
`predominantly synchronous technologies, supporting quick connections. A
`few implementations have included shared objects [Tang and Isaacs 1993],
`but most do not. None of the other features such as one-way drop, context
`regeneration, or threading are supported by videophones.
`
`6Email systems such as Notesmail™, Eudora™, and ccMail™ allow users to program auto-
`matic message classification, but some expertise is required to set this up.
`7One can obviously “share” objects in a rudimentary manner by pointing a video camera at
`them [Gaver et al. 1993; Nardi et al. 1993; 1996; Whittaker 1996a; 1996b], but currently video
`does not provide high enough resolution to allow text to be easily read this way; nor does this
`method allow both partners equal ability to modify the document.
`
`ACM Transactions on Computer-Human Interaction, Vol. 4, No. 2, June 1997.
`
`Ebates Ex. 1013, Page 9 of 32
`Ebates Performance Marketing, Inc. v. International Business Machines, Inc.
`IPR2022-00133
`
`

`

`146
`
`•
`
`Steve Whittaker et al.
`
`Shared workspaces support real-time object sharing, and a number of
`products exist, e.g., Proshare™, ShowMe™, and Deskslate [O’Conaill et al.
`1994], although none is ubiquitous. These systems are intended to support
`rapid connection. There are no explicit features in these systems for context
`regeneration, threading, or one-way drop of information.
`It is clear from the above that no current technologies support all five
`lightweight-interaction features. TeleNotes was designed to rectify this
`situation.
`
`3. DESIGN
`The observational data from face-to-face lightweight interaction enabled us
`to generate a set of five critical lightweight-interaction requirements to
`guide the design of TeleNotes. The aim was to present a unifying user
`interface metaphor for the applications that would support these five
`requirements.
`
`3.1 Presentation Metaphor
`The TeleNotes user interface is designed to be analogous to aspects of
`real-world paper-based interactions, to offer our users a familiar metaphor
`for system interaction. We present electronic equivalents of work-related
`objects such as documents, notes, and folders to offer office workers
`familiar tools with which to manage their lightweight interaction over
`time. TeleNotes is intended to resemble a real-world “messy desktop”
`containing papers which are laid out at specific spatial locations, with
`related information being arranged in stacks [Frohlich 1995; Kidd 1994;
`Malone 1983; Mander et al. 1992]. Information in TeleNotes is therefore
`spatially arranged around the computer desktop, with each stack being
`relevant to a separate ongoing lightweight-communication task. Each stack
`is automatically sorted so that it contains information relating to each
`ongoing conversation such as prior messages and relevant documents.
`These messages and documents are intended to serve as “context holders”
`for each separate intermittent interaction. As with their real-world equiv-
`alents, each stack can be physically arranged and relocated anywhere on
`the desktop by the user. Moving one item in the stack is sufficient to
`relocate the whole stack, given that TeleNotes automatically maintains the
`relations between stack items. The stacks “float” on top of other applica-
`tions, making them highly visible.8 This constant visibility means the
`stacks can remind the user about what communication tasks are currently
`in progress. The fact that stacks are readily accessible allows rapid context
`regeneration of materials related to a specific interaction by simply “open-
`ing” the stack of materials. The stacks are shown in Figures 1 and 2.9
`Figure 1 is a schematic of the computer desktops of two users Steve and
`
`8Some of our users actually found the constant visibility too distracting, a problem we discuss
`later.
`9Our figures are a mix of screen dumps and schematics; we use the latter for clarity of
`depiction.
`
`ACM Transactions on Computer-Human Interaction, Vol. 4, No. 2, June 1997.
`
`Ebates Ex. 1013, Page 10 of 32
`Ebates Performance Marketing, Inc. v. International Business Machines, Inc.
`IPR2022-00133
`
`

`

`TeleNotes: Managing Lightweight Interactions in the Desktop
`
`•
`
`147
`
`Fig. 1. Schematic showing management of three conversations using stacks, for two different
`users’ desktops.
`
`Jerry, who each have three ongoing communications represented by three
`different stacks. Figure 2 is a screen dump from TeleNotes s