New Metaphors in Tangible Desktops

Size: px

Start display at page:

Download "New Metaphors in Tangible Desktops"

Lorraine Johnston
5 years ago
Views:

1 New Metaphors in Tangible Desktops A brief approach Carles Fernàndez Julià Universitat Pompeu Fabra Passeig de Circumval lació, Barcelona chaosct@gmail.com Daniel Gallardo Grassot Universitat Pompeu Fabra Passeig de Circumval lació, Barcelona bestsheep1@gmail.com Maiol Pi Blanqué Universitat Pompeu Fabra Passeig de Circumval lació, Barcelona maiol.pi@gmail.com ABSTRACT This paper provides a brief history and perspective of the actual situation of the tangible interfaces providing desktop metaphors. The authors will analize the desk-to-computer step to provide the accurate view of the current one, the computerto-tabletop step. Analizing also the different approaches at this moment and identifying their possible issues will help to clarify the way we 1 have to take to a tangible computing world. Categories and Subject Descriptors H.5.2 [Information Interfaces and Presentation]: User Interfaces Haptic i/o, Interaction styles, Windowing systems General Terms Tangible Desktop Keywords Tangible interfaces,desktop metaphor 1. THE DESKTOP AS A METAPHOR 1.1 Traditional Desktops Let s talk about traditional desktops, that can be found in every office or home, made of wood or glass, on which the work is performed. Commonly, traditional desktops are formed by several parts. First, the work area where all tools are located, and where actions are taken. Second, folders, drawers or other systems to sort, or to keep the documents. In fact new and used. 1 The tangible interfaces developers This tools in traditional desktops are: pens, pencils, telephone, rulers, rubbers, clocks, etc designed to write, to paint, to communicate. 1.2 Virtual Desktops With the arrival of new technologies, appears the opportunity to put this traditional desktops, phisycal desktops, inside computers, becoming virtual desktops. Like traditionals, virtual desktops are formed by a specific work area, where folders, files and documents are located in. Also several advanced tools are within easy reach of users, providing a big range of actions. The same classic actions can be taken, but more developed. Computers improve traditional actions in sense of efficiency and speedup. 1.3 Traditional to Virtual First of all, what can we obtain from this change? Well, this evolution gives new features to users, maintaining the metaphor of desktop inside the computer. Folder/File system improve the organization of documents, by making a faster and easier search of them. With traditional desktop users can lose documents when there is an excess of them, on the other hand with virtuals, the static organization avoids this kind of problem. Moreover, all new tools that this change provide us are advanced and useful. For example we can use text editors, sound editors, advanced communication tools, printers, etc almost without effort. This change was studied by Jakob Nielsen[6], according to him, first in command-based interfaces we must specify objects and functions (commands) in noun/verb syntax, as Ms- Dos, UNIX. Until the firsts Graphical User Interface (GUI) and windows, icons, menus, pointing devices (WIMP), where objects are represented by icons, and commands in menus. Talking about interaction, GUIs and WIMPs are not good as can be. For example, some interaction techniques are based on complicated sequences (click, select, run, unselect), also some others provide low or no haptic feedback. Mouse can be an example of an efficentless interaction device. Nielsen also studied, how should be the future Tangible User Interfaces (TUI), said that this interfaces should be in some degree free of syntax, and where action and object will be

slower and more complicated. 2. OVERVIEW TROUGH THE DIFFERENT MODEL-VIEW-CONTROLLER SCHEMA 2.

That components are strongly separate, you can have the input tools near and the representation channel may be far ( ie: pc, keyboard and screen can be separated two or three meters; or an electric

2 unified in one token. Also he said TUI will be more efficient because in one time, user can grasp-manipulate, instead WIMP users have to do it in four steps: click-select-manipulate-deselect, making its manipulation slower and more complicated. 2. OVERVIEW TROUGH THE DIFFERENT MODEL-VIEW-CONTROLLER SCHEMA 2.1 User interfaces Typically, users interacts with machines using two communication channels, one for the input data and other for the output, also called control and representation. That components are strongly separate, you can have the input tools near and the representation channel may be far ( ie: pc, keyboard and screen can be separated two or three meters; or an electric guitar, the guitar player can t hear his music if he don t have an speaker near, in a concert they are normally focused to people ) Figure 2: Controller Graphical User Interface Model View Figure 1: User Interface Model View Controller 2.2 Graphical User Interface Lets to analyze the model view controller in the example of GUI S, here you also have two communication channels, an input and an output, but it has some different aspects. Gui s uses desktop metaphor but you can t manipulate desktop elements directly, to do it, you ll need external tools whose function are like a remote control of desktop. As a consequence, the feedback that you can have is too poor ( the remote elements normally don t have any kind of output communication), see table Tangible User Interface The real change appears here with the fusion of the two channels control and representation (figure 3), with this paradigm, you can touch the information, TUI becomes a real digital desktop, then the tangible objects becomes the door between real and virtual world (metaphor of icebergs [4]). But also, with objects, appears some interaction problems, the tangible information is only tangible with objects? like a remote control of the information? no, the objects are only a little portion of the information, digital information is too big to keep it into an object. And how to solve it? there are no way, well, not an only wall to solve it, but this paper try to help to solve it. Figure 3: Tangible User Interface Model View Controller 3. FROM GUI TO TUI After collecting all this information we searched for the current implementations of the tangible desktop. And we almost found nothing. Why can not we find almost anything in relation to this? Is it possible that this path has not been explored yet? Is it a useless research? We found several papers from different researchers that think the same about this questions. Lars Erik Holmquist said[3] Re-thinking the desktop computer has become a largely irrelevant pursuit. It is a battle that has already been both won and lost. We are going to explain this point of view. 3.1 The early problems WIMP systems were developed in the mid-eighties at Xerox Park. Now we have computers a million times more powerful with memories up to GB and, what can we see if we look to a computer? The same boring Windows, Icons, Menus and Pointer devices, our WYSIWYG editors even the Sliders. Nothing has changed from the 80 s. And this is not due to the lack of research about expanding the WIMP paradigm

Action, External Tool (the remote controller) mouse keyboard joysticks, pads, tablets steering wheel, some special joysticks and some electromechanical tools Reaction a handler, a pen or some drawing

ever are the same, a virtual representation of the real world.

uses of that elements in the desktop metaphor world) with it. Table 1: A brief subset of remote control devices but to the fact that every attempt of changing this has end up by failing.

3 Action, External Tool (the remote controller) mouse keyboard joysticks, pads, tablets steering wheel, some special joysticks and some electromechanical tools Reaction a handler, a pen or some drawing tools, it s the representation of your hand in the virtual environment. a writing tool, simply your way to say something in the virtual environment. ever are the same, a virtual representation of the real world. they begins to have an haptic feedback ( like force feedback) but at the moment there are no real application GUI(With it I wanted to say, that at the moment, there are no commercial or noncommercial uses of that elements in the desktop metaphor world) with it. Table 1: A brief subset of remote control devices but to the fact that every attempt of changing this has end up by failing. Isn t this a sign of the saturation of this model? Perhaps it means that we are in the limit of this model. That s why these researchers say that every single attempt of following this direction in tangible interfaces will fail, that it is not possible to change WIMP even through tangible interfaces. 3.2 Actual tangible desktops As an example to illustrate this last opinion we can look through a couple of projects of the tangible desktop: Figure 5: mediablocks 3.3 The proposed approach So perhaps in tangible computing we do not have to search the active interaction between human and computers but the passive one. A way to do this is to not introduce objects in computers but computers in objects (like ubiquitous computing) Figure 4: Panasonic Interactive Table Panasonic Interactive Table 2 attempts to integrate WIMP into a tangible tabletop using some fish tank metaphor where programs (fish food) are executed by fishes and their communications are done by bubbles. Honestly we do not find any sense in here. Far from taking advantage of the capabilities of the tangible tabletop interfaces, the researchers pervert the well-known-to-work WIMP adding them a supposed new metaphor. But this mix of metaphors does not look good either. mediablocks[7] seems to implement the tangible file. In some way it is the natural evolution of the files and folders OS metaphor and a good try. But we have to think that this is not really a new invention. Floppy disks has existed for many years, and now pen drives seem to accomplish the same function than mediablocks. 2 We could not fint the Panasonic Interactive Table reference Figure 6: MediaCup An example of that is MediaCup[1], which is a cup, that has all uses a cup can have, but enhanced. This cup has several sensors on its base that collect and record data about the temperature of coffee, position, acceleration and so on. This information is transferred to a bigger computer that calculates useful data and takes decisions such as the number of coffees that we take along the day; in order to know if we are nervous and therefore to buy less intense coffee. This

Figure 7: reactable An example of this approach is reactable[5] which is a very good approach to the music interaction through a tangible tabletop interface.

4 exemplifies correctly the reasoning of these authors. Another proposed approach is to make orthogonal applications to the desktop. Do not ever use WIMP, and make specific applications for specific purposes. Figure 7: reactable An example of this approach is reactable[5] which is a very good approach to the music interaction through a tangible tabletop interface. There is no pretension to enhance the WIMP model, in fact it does not use it. 3.4 Our own approach We think WIMP and tangible interfaces are not incompatible. We have to take all the useful metaphors and reject all the metaphors introduced by the computers limitations (like mouse pointers). Adapting these metaphors to tangible interfaces is our quest. like icons or symbols), they must be functional without any instructions or reference manual, they only must be functional. The second point of objects design is their relationship way, some objects could be virtually connected, that connection must be light and don t interfere the other relations and the movement area of the other objects, then if objects uses something like wires it will produce a decrease of interaction, but if you uses a virtual relation, is too possible that users lose the concept of that connection. Are objects strongly necessary? No, ever depends of the paradigm that you want to solve, if the application is GUI oriented, the use of the objects will be unnecessary even they could fail the application. 4.2 A brief classifications of tangible objects Objects can be classified in two classes, tangible tools and hybrids[2]. Tangible tools are all objects that contains an strong relationship with a concrete application, like handlers or concrete tools, the representation of that objects are strongly based on his physic representation, then the data that can manipulate is concrete and predefined. Hybrid objects, are the opposite of tangible tools, them have a little physic representation and his function is based on the data that them are assigned at time, this kind of objects can be used for the applications that will be scalable in a future or which requires a great amount of objects. Hybrid tools can change their role dynamically. Returning to the iceberg metaphor, tangible tools would have more ice out of water, and hybrid tools would be the icebergs which only a little part of them are out of the water. Other types of menus, and other concepts for windows and icons, we can emerge some objects to reality: the tangible interfaces strenght is the hybrid real-virtual capacity of the traditionally aonly real or only virtual objects. This is the way. 4. CONCLUSIONS 4.1 A little guide for avoid interaction problems How many tangible objects an application must have? Ever, it depends of the application that will be use them, but the big problem of the objects is their physicality, when you have a lot of objects the interaction experience decrease, then is completely necessary find the balance between objects and interaction (see table 2). low interaction great interaction few objects optimal number of objects difficult interaction a lot of objects Table 2: Relation between objects and interaction Object metaphor and their relationship, how to solve it? The objects would have an strongly reference to cultural or well known references to the real world ( Figure 8: In the first imatge you can view the icberg metaphor of tangible tools, in the second is the metaphore oob hybrid objects 4.3 In summary All we can extract about this are more questions, as consequence of the short live of TUI paradigm, but the only way to try to solve that questions are using user test-based studies and observe all great advantages of the existing technologies(desktop, gui based aplication, console based aplications,..) and try to become them tangibles. 4.4 Future Work Taking advantage of the fact, that two authors of this paper are working with the objects paradigm in our PFC s, we will

SIGGRAPH 98: Proceedings of the 25th annual conference on Computer graphics and interactive techniques, pages 379 386, New York, NY, USA, 1998. ACM Press.

5 SIGGRAPH 98: Proceedings of the 25th annual conference on Computer graphics and interactive techniques, pages , New York, NY, USA, ACM Press. Figure 9: It uses a big object for simulate a drawer, obviously the object is a tangible tool, then is too difficult use it for other applications, but for an only application is too easy for recognize its function comment some examples, this examples are designed for the reactable platform that is based on a multi touch table that can recognize fiducials, the tangible objects. We ve try to forget all wimp ideas that we ve get using actual PC: forget window based limits, the remote control of wimp input,... ; even use simplified objects but finally the great question that we have found is : Whats better, try to make an oriented desktop design or reinvent the digital desktop? 5. REFERENCES [1] H.-W. Gellersen, M. Beigl, and H. Krull. The mediacup: Awareness technology embedded in a everyday object. In HUC 99: Proceedings of the 1st international symposium on Handheld and Ubiquitous Computing, pages , London, UK, Springer-Verlag. [2] T. M. G. Hiroshi ISHII. Tangible user interfaces. In CHI 2006 Workshop Proceedings, page 163, Medford, MA, USA, Department of Computer Science Tufts University. [3] L. E. Holmquist. From human-computer interaction to the use and design of digital artefacts [4] J. H. Israel. Noncommand-based interaction in tangible virtual environments. In Phisicality 2006, page 47, Uk, Lancaster, British HCI group, Equator. [5] S. Jordà;, G. Geiger, M. Alonso, and M. Kaltenbrunner. The reactable: exploring the synergy between live music performance and tabletop tangible interfaces. In TEI 07: Proceedings of the 1st international conference on Tangible and embedded interaction, pages , New York, NY, USA, ACM Press. [6] J. Nielsen. Noncommand user interfaces. Commun. ACM, 36(4):83 99, [7] B. Ullmer, H. Ishii, and D. Glas. mediablocks: physical containers, transports, and controls for online media. In

Figure 11: In the line of desktop metaphor, here you can see an application handler, obviously it s a tangible

different than others, is the metaphor of the windows in GUI based applications.

versatile but the user have to had some information about it before use.

6 Figure 11: In the line of desktop metaphor, here you can see an application handler, obviously it s a tangible tool it s a contradiction for a big application like a desktop environment but in that case, that tool must be different than others, is the metaphor of the windows in GUI based applications. Figure 10: That s an other example of a drawer but, now is using an hybrid object, his functions is more versatile but the user have to had some information about it before use. Is easy to see that the physic object can be used by another application with a different function Figure 12: Finally an example of wiring two applications using an application handler.

Prototyping of Interactive Surfaces

LFE Medieninformatik Anna Tuchina Prototyping of Interactive Surfaces For mixed Physical and Graphical Interactions Medieninformatik Hauptseminar Wintersemester 2009/2010 Prototyping Anna Tuchina - 23.02.2009