For a class lecture on XML I searched for a free text editing tool supporting syntax highlighting for XML, check if XML is well-formed an perform validation (e.g. with DTD). After some research I found out that a plugin for Notepad++ called XMLTools. I was very lucky because I already use Notepad++ for years. The installation of the plugin was no problem and had all the functions I needed for the class. In the first lecture everything worked fine. We checked for well-formed XML, developed some CSS for viewing in the browser and finally started writing a first DTD for a XML file.
Suddenly the students had files which couldn’t be found by the plugin, when they tried to validate their XML. A popup appeared telling us that it was “Unable to load the DTD xyz.dtd”. I examined the files but couln’t find any errors. I searched the net, but although I found some (elder) questions on that error, I didn’t find an answer.
So I took the files home and did some further research (trial and error). First thought was that the coding of the files was wrong, but that didn’t make any change – same error message.
Finally, it was more than one source of error. This told me once again, that error messages often don’t describe what really was the error (and that there should be different error messages for different errors!).
- Every syntax error in the DTD causes the “Unable to load” error. This means, if only one bracket is missing, a wrong character or similar small errors is contained in the DTD, the popup is displayed and no validation is possible.
- The second source of error actually was the coding: UTF-8 coded DTD-files can’t be loaded by the plugin. We were only able to validate XML files with ANSI coded DTDs (without error!)
Isn’t it surprisingly, how easy it can be to experiment with human perception without any tools? Here is a very simple one for binocular disparity.
Because of the space between our two eyes we perceives two distinct images. The difference of the images on our retinas can be easily recognized with the thumb experiment: You only have to put forth your hand, thumbs up. Now, if you first close the right eye and then the left one, the thumb seems to be at different places in comparison to the background (see Figure).
The very popular stereograms were formerly used for experiments on disparity. You stare at two distinctive images that have a relatively small distance from your eyes and try to look through them. When you have finally reached a rigid view, you get a 3D effect. You thereby switch off the convergence of the eyes.
Finally it is done – my doctoral thesis has been published:
The dissertation called “Der Interaktionsprozess bei Tangible Interaction – Entwicklung und Analyse eines Interaktionsmodells – Richtlinien zur Verwendung beim Design” is avaliable in Verlag Dr. Hut, ISBN 978-3-8439-0836-8. It is only available in german by now.
Being a subarea of Human Computer Interaction, Tangible Interaction is a multidisciplinary area. Since the first prototypes of systems with interfaces that physically represent digital information in reality and therefor provide an easier way for manipulation, this field of application grew explosively.
New technologies allow sensors to become smaller and cheaper and monitors to become bigger. This means new possibilities for interaction between human and computer arise.
This work addresses the design of a generic model for the interaction process of Tangible Interaction.
By means of this model the process can be examined with respect to critical issues. These findings on the other hand give an approach to optimize the design of Tangible Interaction systems in providing especially focused design guidelines.
Although every university has it’s own opinion survey on classes and lessons, I also appreciate reading the entries of students in my class profiles on meinProf.de, a platform where students can rate different aspects of my teaching like fairness, assistance, material, comprehensibility, fun and interest. Especially the free text field is very interesting, because often the students give some good clues on what could be done better.
Here are the links to my profiles (they are both in german language).