tag:blogger.com,1999:blog-1701128444935652651.post1226239314045240491..comments2009-02-25T18:01:49.469-08:00Jonathan Brillhttp://www.blogger.com/profile/02653184610453479546noreply@blogger.comBlogger1125tag:blogger.com,1999:blog-1701128444935652651.post-53366196694570318122009-02-12T17:50:00.000-08:002009-02-12T17:50:00.000-08:00Overall comment:<BR/><BR/>This book is what I have been waiting for so long time. Even with the first chapter only, it is really helpful for me to get myself organized regarding gestural interfaces. I have two wish lists for the first chapter.<BR/>1) Explicit classification of gestures: In my opinion, types of gestures may include touchscreen gestures, body motion gestures, 3D hand/finger gestures, pen (tool ?)-based gestures, etc.<BR/>2) Relationship between other HCI paradigms: There have been many HCI paradigms such as ubiquitous computing, pervasive computing, organic user interfaces, and so forth. I think the common factor of those paradigms (including gestural interfaces) are the maximum usage of sensor technologies. And the gestural interfaces are the first of this kind that come into the real life while others still live in labs. I think that the position of gestural interfaces in HCI field can be clearly shown by describing this relationship.<BR/><BR/>Detailed comments:<BR/><BR/>p. xiv: His motivation to write the book is exactly same with the reason why I made my blog. I'm very happy to get the first book to try to collect and organize various aspects of gestural interaction.<BR/><BR/>p. 15: Let me introduce my definition of touch event. For me, the following definitions have been useful when I implement touch interfaces.<BR/><BR/> - Touch Event: An event that is produced during or after a user performs a touch gesture. An abbreviation of Touch Gesture Event.<BR/>Event: A message notifying that internal states of a system is changed.<BR/>Touch Gesture: A gesture that a user performs on a touchscreen or an interactive surface.<BR/><BR/>p. 20: It is said that 100 ms or less is ideal for instantaneous responses. I think it would be better to replace 100 ms with 10 ms. When I used touch interfaces running at 30 ~ 40 hz (25 ~ 33 ms) systems, I got latency problem. From my experience, I prefer to use 100 hz (10 ms) system. 100 ms seems to be slow for most gestural interaction systems. When I analyzed handwriting motions, their peak frequency was around 20 - 30 Hz. That is another reason I usually stick to 100 Hz sampling rate.<BR/><BR/>p. 25: For the paragraph "Number of participants", I think MERL's DiamondTouch could be a better example of detecting multiple users. The sensor system of DiamondTouch gives different electrical signals for different users while that of surface doesn't.W. Changhttp://uuook.chang.googlepages.comnoreply@blogger.com