I spent the last two days to find a way for automatic linebreaking in java. The simplest solution would have been to use the greedy algorithm. You get far better results using the Knuth & Plass linebreaking algorithm, which has been developed for TeX by Donald Knuth. Looking for a java implementation I first found two frameworks folio and apache fop that implement the algorithm but also involve a couple of classes and complex possibilities for import and export. So I decided to use the implementation of Simon Pepper wich can be found here.
I modified his test class to not read the text and setup using xml but instead just break given strings. I also had to change the calculation of the linewidth to use the pixelwidth of the characters dependend on a certain font. After finishing the linebreaking I was looking for a good solution for hyphenation. I found hyfo which is part of the already mentioned folio package and can be downloaded from sourceforge. Hyfo is using the hyphenation ruleset of latex that are available for different languages. The final result can be found at my google code repository under trunk cc.creativecomputing.util.
this is one of my first sketches created with creative computing. it is based on steering behaviors and perlin noise . there are 5 agents moving controlled with different behaviors. the first one has applied a wanderer behavior. the other four have applied wanderer and leader follow. only the 4 following agents are drawn. the agents have tail that is slightly modulated with a perlin noise.
Animation created with creative computing. This sketch is inspired by calicraft from ricard marxer. It is based on steering behaviors and font outlines. there are three swarms of agents that have applied flocking and path following. after a certain time every swarm is set free and than set to follow the path of another letter. creative computing gives the possibility to create font outlines, which can than be used as input for the pathfollow behavior.
For quiet a while I have been working on my own framework for interactive graphic applications. It started as a new OPENGL renderer for processing. During my diploma I made some changes on the processing code to optimize the drawing process. I then realized that it would make more sense to not only create another renderer but also adjust the processing code to get real benefit from OPENGL. But still I wanted to keep things as simple as possible so I wrapped a lot of OPENGL calls in shorter functions for simpler and faster coding. Besides this changes I also put all the utility functions of procesing to different classes. Over the last year I made a lot of changes to the original processing code texture handling is now based in the Texture classes of jogl. I also added a lot of code from different processing libraries and forum entries so you find fog and camera as well as different other helpful things added to the framework.
I also changed most of my libraries to make them work with creative computing, this is one of the reason why haven’t provided updates for the processing versions for a long time. Although I hope to change that. There are still many things to do especially documenting. But I started to put the code online on google code.
Idea 1 Capture of movement
The human body is one of the most chosen themes in sculpturing, capture the movement using a still is the most difficult task. For 3D Animation it is a common technique to use camera tracking and markers on persons to capture there movement. The idea is to use this data for a sculpture of the movement. Instead of expensive camera tracking the aim would be to equip the tracked person with a number of acceleration sensors and than extrapolate the movement out of this data. Especially tracking of sports like gymnastics could result in interesting images.Another idea would be to track the movement of a ball of a basketball or other ball game.

Idea 2 News
Here the idea is to show the relation between countries statistics and there relevance for the public media. So for example death statistics of every country are looked for the number of killed people and than set in relation to the number of articles found in google news.
Another Idea would be to use the possibilities of Archive research to search for general developments in use. Possible questions could be:
How is the relation of different categories region, world, politics, sport, business etc.
How is the relation between text and image?
How long are articles?
What can be found on the mainpage?

Idea 3 Who Counts?
One big problem with information visualization is that you can only show what exists. In less developed countries statistics are often not complete or even missing. But also with existing you always have to question it. For example for american census data some people claim that poorer parts of the society are excluded from the statistics to make it more positive.

One idea would be to find new ways in aquiring data for example one could create cheap devices that check the noiselevel or airpolution and store them together with the location. Later the data is uploaded to the web and can be shared with the public.
After a long time absence, I decided to give my blog a new chance. Today I passed my oral examination on the University of Applied Sciences and finnished my studies in Computer Science in Media. In my thesis I developed a visualization of the worlds air traffic. I used pdf flight schedules from the biggest the airline aliances as datasource and mapped the flights based on their start and destination airport.

The image shows more than 90.000 flights of selected airlines for August 2007. More information on this will follow on an extra project page.

In October I gave a one week workshop at the HfG Karlsruhe. Primarily designed for non-programmers, it presented an introduction to the syntax and semantics of the Processing programming language, as well as an introduction to computational aesthetics. By creating computer applications, instead of just using finnished software, the workshop should give a deeper understanding of the essential possibilities of computation. Go to the workshop page for more information.
procontroll was updated to work with the new jinput version. it is now running in a seperate thread to allow more exact event handling. with the new plug method you can register methods to be called on button pressed and released events. I also added a new cooliehat class to correctly detect this kind of button. Get more information at the procontroll site.
I have updated the proXML lib. In the new version xml files can be loaded in a seperate thread. This is usefull to load xml from other urls without causing your sketch to wait.
There are two new methods getText() and getName() on XMLElement, infact they do the same as getElement() but maybe it is less confusing. There is also a new method for removing child nodes from an Element. And there are new methods for accessing childnodes and siblings, so you can get the firstChild, lastChild, nextSibling and preoviousSibling.
I also fixed some bugs with parsing xml.

