I have followed the NSA leaks very closely from the beginning. When the leaks started and Snowden firstly revealed himself I had been in Berlin and some tech-savy people involved in the Bitcoin scene told me about the leaks. Following the leaks closely, watching Snowdens first video and the upcoming revelations have definitely influenced me. I curiosly read Glenn Greenwalds book and watched the Laura Poitras movie “Citizenfour” soon after they were available.
In the summer of 2014 I got the possibility of creating a tool which, in my opinion, was a reasonable and necessary step in the direction of making the supervision of authorities more accessible. The german organization netzwerk recherche is a journalists association with a focus on investigative journalism. German journalists have certain extended rights to question official institutions (such as secret services) on data which has been saved about them. This is meant to strengthen their journalistic role within a democracy. These rights, however, are seldomly invoked. Concerning the secret services, a reason could be that there are actually 20 (sic!) intelligence agencies/secret services in Germany — one for each state and four for the entire country. Each service has different requirements for answering requests — one might e.g. require a copy of the passport while another might demand other documents. Additionaly each service has a different address, which is not always easy to find. So to ease these hurdles the idea came about of developing a simple PDF generator, where one could just click the agencies one wants to inquire. A PDF containing the necessary inquiry text, attachment information and the address would then be generated.
A special requirement for such a tool was that the PDF document generation had to happen on the client-side — no server should hold any state. It must not be possible for any person having access to the server to monitor who wants to take use of ones informative rights. Also one should not have to trust the organization, instead one should have the possibility of downloading and deploying the generator himself. Making the source code available (as free software) was a natural conclusion.
I have created this tool and it has now been deployd since July or so. I just didn’t get around to do a proper writeup.
The public instance is available at the netzwerkrecherche.org website. The source code is available via GitHub (under the MIT license).
Das netzwerk recherche ruft Journalisten auf, bei den Geheimdiensten anzufragen, ob diese Daten über sie gespeichert haben. Um das zu vereinfachen, stellen wir einen Generator für die entsprechenden Anträge bereit.
Ziel dieses Projektes ist, in Zeiten der zunehmenden Massenüberwachung den Diensten zu zeigen, dass ihr Handeln von der Öffentlichkeit kritisch beobachtet wird. Die Aufmerksamkeit für das Problemfeld Geheimdienste im demokratischen Rechtsstaat muss erhöht werden.
Insbesondere für investigativ recherchierende Journalisten ist eine Überwachung durch Geheimdienste und eine damit einhergehende Ausforschung ihrer Informanten und Kontakte nicht hinnehmbar.
Anlass des Projektes ist der Fall Andrea Röpke. Beim niedersächsischen Verfassungsschutz wurden offensichtlich rechtswidrig Daten über sie gesammelt. Als sie einen Antrag auf Aktenauskunft stellte, vernichtete die Behörde ihre Akte und behauptete, es gäbe keine Akte über sie. Erst ein Machtwechsel in Hannover offenbarte die Vertuschung – die politische Aufarbeitung läuft bis heute.
The Institute of Distributed Systems at the University Ulm runs the course “Practical IT-Secutiry” this semester. To me this was especially interesting because of the examination modalities: one does not have to take an exam, instead each student has to prepare and hold a lecture and accompanying assignments on a certain topic. I decided to dive deeper into Web Security and chose for Cross-Site Request Forgery (CSRF) attacks.
The presentation can be found online here. The preparation document for students was distributed one week in advance to the two-hour assignments (pdf). Assignments were based on the Metasploitable framework, the Damn Vulnerable Web App and TWiki. Additionally, I wrote some intentionally vulnerable PHP scripts with increasing levels of security.
All of this material (*.tex, *.html, *.pdf, etc.) and solutions for the assignments can be found in my talks GitHub repository as well.
The german weekly newspaper “DIE ZEIT” has an API available. This means it is easily possible for developers to use a lot of their data. Since they have made access to the data of nearly 400.000 articles since 1945 possible this is quite interesting (access to full texts is sadly missing, but a lot of other stuff is available). This post is about some of the interesting things I found whilst exploring the API.
My initial idea was to visualize how the ratio of articles with anglicisms evolved over time. At the moment this is too complex a project, due to the fact that getting the necessary data via the current API is difficult. However, I made some other interesting findings along the way.
The Wiktionary project provides a list of anglicisms (around 960 words) which I parsed out and used to search for articles concerning these words. This gave a list of how many matching articles on this word had been written each year since 1945. I also made an empty search to find out how many articles were created in total each year. These numbers could then be used to calculate the percentage of articles with anglicisms in each year.
Not all of the words provided interesting results but here is selection of some interesting ones. Please be aware that the statistics show a zoomed-in range. This is not a scale of 0-100%!
One should be very careful to interpret reasons for the peak just by looking at the visual representation. A potential reason might be the Gulf War in 1990–91 (the german translation is: “Golfkrieg”). Other causes worth investigating could be successes of german golf athletes or events around the VW Golf automobile.
Potential reasons for the peaks could be: in 1985 the Sinking of the Rainbow Warrior, in 1995 the Brent Spar protests and in 2010 the Deepwater Horizon.
The peak in 1987 could relate to the increased media coverage on aids. Also in 1987 the Institute for German Language (Gesellschaft für deutsche Sprache) chose “aids” has as the word of the year.
The peak in 1970 is most interesting to me, a potential cause could be the movement of 1968.
I have made the code used to gather the data and build the visualizations available under the MIT license via GitHub.
Valerie is currently presenting on the ongoing exhibition “Kunst oder Kitsch?” in Bad Schussenried (13. April – 22. June 2014). The idea was to let different artists (eighteen in total) explore where the separation between art and kitsch lies. What is still art and what is kitsch? Where does the line lie?
I liked the exhibition a lot! The topic is really interesting and made me curious immediately. Some artists made artworks which run really hard on the borderline between art and kitsch and it is a lot of fun walking through the exhibition and discussing about specific creations and their classifications.
Some artists had the idea to complete the exhibition with a — not that seriously meant — interactive installation where visitors have the possibility to rate the exhibition: does it rather present kitsch or art? They asked me if I would like to build such a thingy. I started with the software and after a while Leo joined and developed the electronic hardware part: two large footswitches need to be pressed in order to vote. Valerie also greatly helped in building the funfair-like wooden base for the “Kitschometer”.
It basically works like this: you press one of the buttons to vote if the exhibition rather presents art or kitsch in your opinion. A random, fancy song will then be played and a tachometer will hit out on either art or kitsch, depending on how the previous voters have rated. The exhibition runs for three months and I think it’s a nice gimmick.
The code and technical documentation is up on GitHub. The installation could surely have been build technically more elegant, but in this case a pragmatic approach was taken due to time constraints. The footswitches are borrowed from the workshop at the University Ulm. This is exactly how a university should be! It should enable people to just do stuff and not constrain them in realizing ideas!
I would definitely like to create more installations.
Update:
The exhibition ran for about three months, about 3.800 visitors attended. The final score of the installation was: art – 882, kitsch – 772.
