Evaluation and Final Thoughts

According to [19,3,17] exchanging, editing or visualizing media using a traditional GUI approach is not always as easy as it may appear. The variety of media, the complexity of exchanging it amongst the various devices, the size of the data (in terms of duration, volume or dimension) are some key factors for the clumsiness of purely graphical approach towards media manipulating. As a consequence, inquiring for new ways of achieving these tasks does not seem void.
Interstingly enough, while all three articles present concepts for new solutions their empirical results emphasize how impressive approaches can be reduced to useless just by failing to solve some (apparently) minor issues.
Hence, while the idea of the game board in [3] seems of great interest, the technical limitations introducing unreliability and variable latency reduces the user acceptance by an important factor. Thus, by their work on logJam Cohen, Withgott and Piernot do not only show the benefits of tangible artifacts but also emphasize two keys for a usable TUI: reliability and responsiveness. Towards the seeking of a valuable metrics for TUI evaluation these two terms seem to be amongst the most important.
Logjam failed to fulfill them because the identification system was lacking in flexibility. Indeed, the user was supposed to place the blocks on the board in a specific manner in order to initiate a galvanic contact between the token and the main bus. This approach is in contradiction with the real physical world where placing things usually do not require for us to place them in an exact manner.
Technology advances are helping worthy ideas to meet the requirements. Going in the direction of real world ergonomics, the contact less identification scheme (using RFID) for the tiles in [17] seems to be more reliable and responsive.
Flexibility is not only important in terms of ease-of-use but also in terms of use-cases. A modern TUI should not be bound to a single use but ought to offer an interface for various manipulations. The DataTiles and mediaBlocks projects represent a step towards this aim. However, for such generically projects to get out of the laboratory some so called ``killer app''<sup>10</sup> are required. Hence, while mediaBlocks and DataTiles provide an impressive framework for media manipulation and interchange, both projects somehow fail to reveal totally convincing applications.
Nevertheless, all three projects exhibit many benefits from using tangible objects as interfaces. First of all logJam emphasizes some idiosyncratic uses of the interfaces such as the way each user places the tokens to fit his own perception of organization. This fact is the direct consequence of a great benefit of TUIs over GUIs as exposed in [19]:
[...] the GUI metaphor appeared unable to generalize across the potential design space of tangible user interfaces
or in [3]:
[...] TUIs allow people to take advantage of all the degrees of freedom available in the physical world.
TUIs introduce the malleability of real world objects to control the digital world. This malleability coupled with flexible composition schemes enabled the guinea-pigs of the DataTiles project to create non-anticipated combinations close to real world ergonomics.
Moreover, the undeniable benefit from physical contact is emphasized by the three projects as in [3]:
[...] there was an unmistakable sense of physical contact [...]. Such physical sensations do not exist on the screen.
Besides physical contact, physical exercise introduced by tangible user interfaces is also of great value. As an example its benefits in the learning process are distinct (see [10]).
These facts alone are enough to justify the TUI researches in the media domain. However, the technologies still need to evolve towards reliability, flexibility and responsiveness. On their side, users have to overcome the ``de facto'' ergonomics introduced by years of GUIs monopoly in order to appreciate the real benefits of tangible user interfaces for manipulating media content.