CHI 2006 · Work-in-Progress April 22-27, 2006 · Montréal, Québec, Canada Swordfish: User Tailored Workspaces in Multi-Display Environments bindings, each user can choose their own display connections and create a personalized MDE. This approach also helps manage changes in the environment as devices enter, move, or leave. Jim Wallace Vicki Ha Dalhousie University Dalhousie University 6050 University Ave. 6050 University Ave. Halifax, NS, Canada Halifax, NS, Canada [email protected] [email protected] Ryder Ziola Kori Inkpen Dalhousie University Dalhousie University 6050 University Ave. 6050 University Ave. Halifax, NS, Canada Halifax, NS, Canada Introduction [email protected] [email protected] The tight coupling between monitors and processors characteristic of personal computers over the past two decades is eroding. It is becoming more common for people to utilize multiple display environments (MDEs) for their desktop work [7]. In addition, conference rooms and public venues are starting to provide access to a myriad of public displays. As displays become ubiquitous in our environment, we need to think carefully about how information is managed across these displays and how users (in particular, collocated collaborators) will interact with these displays. ACM Classification Keywords Abstract This paper presents a novel interaction metaphor for Multiple Display Environments (MDEs) called lightweight personal bindings. This approach enables users to easily bind edges from one display to another and move seamlessly between displays. The goal of this work is to support collocated collaboration in a dynamic multi-display environment while accommodating users’ personal preferences. With lightweight personal Copyright is held by the author/owner(s). CHI 2006, April 22–27, 2006, Montréal, Québec, Canada. ACM 1-59593-298-4/06/0004. H.5.3 Group and Organization Interfaces: Collaborative Computing. Recent research in the area of MDEs has stressed the importance of connecting personal devices (i.e. laptops) with shared displays to support collaboration [12]. This not only necessitates an underlying architecture that allows multiple, heterogeneous devices to be 1487 CHI 2006 · Work-in-Progress April 22-27, 2006 · Montréal, Québec, Canada personal bindings. Finally, we reflect on initial usage of Swordfish and directions for future work. Navigation in Multi-Display Environments The benefits of multi-display systems for single users have been well established in the literature [5, 7]. Research on collaborative multi-display systems has stemmed from previous work on public and semi-public displays [7, 8] as well as the interconnection of multiple heterogeneous devices [3, 10, 11]. Figure 1. A sample MDE with bindings between displays connected, it also requires that users understand how to work seamlessly in this larger virtual workspace. Given that many people choose to configure their desktop environments differently [7], it is likely that there will also be differences in how people want to configure their virtual workspace environment. We therefore need to examine how collaborative MDEs are organized and how they are interpreted by users. The goal of our research is to develop a lightweight approach that allows users to easily create and then navigate across multiple displays through virtual connections. Lightweight personal bindings are user defined connections from the edge of one display to the edge of another display in a MDE. These connections can be created and altered quickly, allowing users to define personalized workspaces that support their individual preferences and perspective of the space. This paper first discusses related work in the area of MDEs. We then describe Swordfish, an application we developed to explore the concept of lightweight Various approaches have been investigated to support navigation between displays within an MDE. PointRight [9] and TeamSpot [15] utilize a spatial metaphor to encode connections between displays in which the user moves their cursor off the side of one screen and it appears at the border of the display adjacent to it. Another approach is the use of labels for each display in the environment. For example, Mighty Mouse [3] maintains a list of all displays in the environment and users can select the display they want to control by clicking on an icon associated with that display. More recently, Biehl and Bailey [2] developed a hybrid approach that utilizes an iconic representation in which the icons are laid out to reflect the physical position of the displays in the room. Head tracking approaches have also been applied to control cursors in multidisplay environments [1]. The use of head tracking allows users to effortlessly transfer control to any display in their line of sight, essentially creating fully flexible bindings. Several open source and commercial systems are available for network control of multiple systems (e.g. Synergy [13] and Desktop Rover [4]). However, these 1488 CHI 2006 · Work-in-Progress April 22-27, 2006 · Montréal, Québec, Canada systems are primarily designed for individual users connecting to multiple machines and only allow one person to utilize a display at a time. A key challenge in designing collaborative MDEs is providing support for individual as well as group processes. It is naïve to think that every user will view the MDE in the same way and want to interact with it in the same manner. Just as people like to configure how their windows are laid out on their desktop, people will likely have preferences about the layout of their multidisplay workspace, dictating how they move between different displays [7]. Lightweight Personal bindings The idea of lightweight personal bindings is that users dynamically ‘bind’ specific edges on their personal device to other displays in the MDE (i.e. laptop to various public displays) to provide a virtual connection between the displays (see Figure 1). Once a binding has been created, moving the cursor off the screen in the direction of the binding moves the cursor to the connected display. Any display can have bindings to several other displays and multiple users can bind to, and actively interact with, a display simultaneously. The main advantage of this approach is that users can freely alter the shape of their workspace to meet their personal needs. In addition, users are likely more cognisant of the virtual connections between displays since they have chosen which bindings to create and have manually instantiated them. Only displays that are being used require bindings, thereby reducing the complexity of a user’s workspace. User created bindings have previously been investigated for tablet PCs. For example, ConnecTables placed in close proximity to each other automatically form a connection between the adjacent edges [14]. Stitching [6] is another technique that allows users to connect mobile devices on the fly. This approach allows users to create bindings using a pen-based gesture across the devices. Our lightweight personal bindings extend these concepts to more diverse environments and permit more connections per device. Swordfish Testbed As part of this work, we developed Swordfish (Spontaneous Workspace Organization Designed For Interactive SHared environments), a testbed application to explore lightweight personal bindings. It was developed in C# using Microsoft .NET. The ability to create arbitrary bindings between displays in a MDE is available in other research and commercial systems [9,15], however, development of our own suite enabled us to explore many underlying components of display management systems. The Swordfish framework is composed of three basic elements: network hosts, user input devices, and displays. Each network host has a number of different input devices and displays associated with it, and maintains a database of all displays and associated security permissions on the network. The network hosts communicate with each other using a peer-to-peer protocol which automatically detects any other available public displays on the subnet. This approach allows users to view all available displays and easily create a binding to any display. 1489 CHI 2006 · Work-in-Progress April 22-27, 2006 · Montréal, Québec, Canada Swordfish renders its own cursors, enabling multiple users to each have a fully functional cursor on the same screen. When a user’s cursor leaves their local screen, input is forwarded directly to the receiving host via a UDP stream of mouse and keyboard events. The receiving host draws the cursors and places the events on the system’s global event queue. The remote host then treats these events as if they were local input. Swordfish’s ability to simulate a multi-user environment is limited, of course, by the input restrictions of the underlying OS. Specifically, only one open window can have mouse and keyboard focus and only one user can drag at a time. Because it is intended for collaboration among small groups, these limitations can be largely mitigated through social mediation. This compromise allows Swordfish to occupy the middle ground between a single user multi-display environment and true groupware, without the significant overhead of either option. As well, custom applications can be developed to handle mouse input from multiple users to achieve the full benefit of a multi-user system to iteratively design, explore, and evaluate various interface components for lightweight personal bindings. To date, we have utilized this sandbox application to examine various interface approaches for managing (creating and modifying) and visualizing lightweight personal bindings. Managing Bindings We have investigated three main approaches for managing lightweight personal bindings: menu, direct manipulation and graph based interactions. All three approaches use standard interaction techniques which make them easy and intuitive to use. The menu-based interface provides users with a dropdown contextual menu for the creation of bindings. The menu is obtained by clicking on the border of the screen where the binding will originate and provides both a textual label and a thumbnail image for each available display (see Figure 2a). Bindings are created by selecting the menu item corresponding to the target screen. The same technique is used to de-select a screen and delete the binding. One of the main purposes of Swordfish is to enable us Figure 2. a) A border interaction technique b) a Palette-based interaction technique and c) a conceptual map-based interaction technique 1490 CHI 2006 · Work-in-Progress April 22-27, 2006 · Montréal, Québec, Canada The direct manipulation approach uses a drag-and-drop (or drag-and-throw) technique with thumbnail images of each display organized in a palette for quick and easy access (see Figure 2b). Users select the desired display from the pool of thumbnail images and drag (or throw) the representative icon to the border of the screen where the binding is to be created. Bindings are deleted by clicking on the edge of the display where the binding exists and dragging the icon off towards the palette. If spatial information is important the icons can be laid out in a manner similar to ARIS [2], however this approach requires that the location of the user’s personal device within the MDE is known in order to create an appropriate map. Our third technique manages bindings using a conceptual map of the user’s workspace. This ‘graph’ view displays a thumbnail for each connected display with links representing the bindings between displays (see Figure 2c). All remaining available displays are shown on an adjacent panel. To create a binding users click to draw an edge between two thumbnails. To delete a binding, users grab and drag the edge representing a binding off the thumbnails. Visualizing Bindings Once bindings have been created, moving seamlessly between displays is possible if users’ remember the bindings (i.e. which edges of which displays are bound together). Visualization techniques can be used to help recall bindings. The most straightforward approach is to place a small, coloured border at the edge of the screen (colours can be chosen to uniquely identify particular screens in the environment). Optionally the border may also contain a thumbnail image or label as an additional reminder of which display is connected to the binding. The conceptual map approach to managing bindings also provides a visualization of the bindings. The map is a graph illustrating the virtual workspace (i.e. all of the connected displays) where the user’s current active display is enlarged and positioned in the centre of the visualization. The nodes in the graph are continually updated as the user moves throughout the workspace to reflect the current position of the user’s cursor. This approach provides a reminder of which display the user’s cursor is currently positioned on and all adjacent, connected displays. Visualizing bindings becomes especially important when several bindings are connected to one edge of a display. The visualization must show where on an edge the cursor must cross to move to the connected display. In addition, users must be able to easily manipulate these bindings, changing their size, shape or position. This optional customization supports personal preference which can aid recall of the binding. Subtle tradeoffs exist between visual effectiveness and visual overload. Borders can distract from the desktop if made too wide, or if too small can be difficult to acquire. Palettes and maps may be effective, but require screen real estate. Creating a lightweight interface that is easy to interact with requires the designer to walk a fine line between creating a usable interface and imposing on the user’s day to day work. Conclusion and Future Work The concept of lightweight personal bindings adds a new dimension to previous approaches to collaborative MDEs. Similar to previous work in this area, our approach enables multiple users to simultaneously share and interact with communal displays. 1491 CHI 2006 · Work-in-Progress April 22-27, 2006 · Montréal, Québec, Canada Additionally, users can create a set of personal bindings, thereby defining a distinctive virtual workspace that integrates their personal devices with shared resources and reflects their perception of the space. A further benefit of this approach is that it effectively manages dynamic MDEs where devices and people enter, interact with and leave the environment. Through iterative prototyping and in-lab usage, the Swordfish testbed has provided us with a great deal of insight on the design of MDEs. We are currently incorporating the ability to move artifacts between displays into Swordfish and examining more in-context usage by outside users. This will enable us to validate the necessity and benefit of lightweight personal bindings. In addition, we are further examining how to effectively visualize bindings in a MDE. Acknowledgements We would like to thank NSERC and NECTAR for supporting this research, and members of the EDGE Lab for their assistance with this project. References [1] Ashdown, M., Oka, K. and Sato, Y. Combining Head Tracking and Mouse Input for a Gui on Multiple Monitors. Ext. Abstracts CHI 2005 (2005), 1188-1191. [2] Biehl, J.T. and Bailey, B.P. Aris: An Interface for Application Relocation in an Interactive Space. In Proc. GI 2004, Canadian Human-Computer Communications Society (2004), 107-116. [3] Booth, K.S., Fisher, B.D., Lin, C.J.R. and Argue, R. The "Mighty Mouse" Multi-Screen Collaboration Tool. In Proc. UIST 2002, ACM Press (2002), 209-212. [4] DesktopRover. http://www.neslosoftware.com/.Accessed Sept 2005. [5] Grudin, J. Partitioning Digital Worlds: Focal and Peripheral Awareness in Multiple Monitor Use. In Proc. SIGCHI 2001, ACM Press (2001), 458-465. [6] Hinckley, K., Ramos, G., Guimbretiere, F., Baudisch, P. and Smith, M. Stitching: Pen Gestures That Span Multiple Displays. In Proc. Adv. Visual Interfaces 2004, ACM Press (2004), 23-31. [7] Hutchings, D.R., Smith, G., Meyers, B., Czerwinski, M. and Robertson, G. Display Space Usage and Window Management Operation Comparisons between Single Monitor and Multiple Monitor Users. In Proc. Adv. Visual Interfaces 2004, ACM Press (2004), 32-39. [8] Izadi, S., Brignull, H., Rodden, T., Rogers, Y. and Underwood, M. Dynamo: A Public Interactive Surface Supporting the Cooperative Sharing and Exchange of Media. In Proc. UIST 2003, ACM Press (2003), 159- 168. [9] Johanson, B., Hutchins, G., Winograd, T. and Stone, M. Pointright: Experience with Flexible Input Redirection in Interactive Workspaces. In Proc. UIST 2002, ACM Press (2002), 227-234. [10] Paek, T., Agrawala, M., Basu, S., Drucker, S., Kristjansson, T., Logan, R., Toyama, K. and Wilson, A. Toward Universal Mobile Interaction for Shared Displays. In Proc. CSCW 2004, ACM Press (2004), 266-269. [11] Rekimoto, J. and Saitoh, M. Augmented Surfaces: A Spatially Continuous Work Space for Hybrid Computing Environments. In Proc. CHI 1999, ACM Press (1999), 378385. [12] Shen, C., Everitt, K., and Ryall, K. Ubitable: Impromptu Face-to-Face Collaboration on Horizontal Interactive Surfaces. In Proc. UbiComp 2003 (2003), 281288. [13] Synergy. http://synergy2.sourceforge.net/index.html. Accessed Sept 2005. [14] Tandler, P, Prante, T, Müller-Tomfelde, C, Streitz, N and Steinmetz, R Connectables: Dynamic Coupling of Displays for the Flexible Creation of Shared Workspaces. In Proc. UIST 2001, ACM Press(200 1), 20. [15] TeamSpot. http://tidebreak.com. Acccessed January 2006. 1492
© Copyright 2026 Paperzz