The Mad Philosopher

Hey, check out this screen shot, paying close attention to the banner ad:

I was browsing Slashdot one day, and I looked up at the banner ad and to my utter shock I saw the top of my own website! That sure threw me for a loop. What kind of practical joke was this? I mean, what was my website doing as a banner ad on Slashdot? I certainly didn’t order one.

It took me about 30 seconds to figure out what was going on. Basically, this surprise was an artifact of a strategy for eliminating banner ads, combined with the coincidence that my proxy server and web server were on the same machine. You see, I (or rather Bruce) had set the /etc/hosts file on my server to redirect requests for certain ad sites to point to 127.0.0.1, the local machine. So when the Slashdot page requested an ad from one of these sites, my desktop computer, via the proxy, looked for the ad on the machine that was hosting my website. Well, the top page for that web server just happens to be the Mad Philosopher website. And so that’s what got drawn.

I’m still surprised that this would occur. I didn’t know that the ad boxes could handle an entire HTML page. I just figured they would be image files or something. But, it happened.

I was contacted by a fellow Canadian asking me to do an interview on the OTAS project. He is a freelance writer who is hoping to pitch our story to a Canadian magazine, plus he is considering writing about this project for his Master’s thesis in Philosophy and Public Policy. Interesting, eh?

So I decided to post his questions and my responses to my weblog. These are my thoughts on OTAS, plus some of my background.

(1) How would you describe this project? (Maybe both in complex, and then really simple terms). What will it do (and how)?

The name of the project provides a description of what it is: Open Tsunami Alert System (OTAS). Basically, we want to set up a system that will alert people who would be affected by a given tsunami. This system will involve tsunami detection and then the dissemination of an alert by various means to areas that are in danger.

The “Open” part of the project means several things. First, the project design and implementation are being done by volunteers through a collaborative effort over the Internet. Second, while not intending to replace existing and up-coming governmental and inter-governmental alert systems, the open philosophy of OTAS is key to its existence and usefulness. That is, it is (or will be) designed to meet the specific need of the people on the beaches of the world who actually need to receive a tsunami alert. As Robert Cringely pointed out in his “Political Will” column, government bureaucrats are all too good at talking to other bureaucrats and not to the people they are intended to serve. The open nature of OTAS will be able to offer a reliable and timely tsunami alert service to anyone that wants it, regardless of political borders or bureaucratic inertia.

To further discuss the meaning of the term “Open”, let me quote Charlie from the project’s FAQ:

• First, it will be “open source”: the copyright will be retained, but explicit permission to copy for free will be granted to everyone. This means anyone who wants can get access to it.
• Second, it will be open in the sense that we hope to let nearly anyone set up an OTAS “client” to receive alerts, and tsunami predictors to generate alerts from seismic data.
• Third, it will be open in the sense that our algorithms and protocols will be public and documented, so others can build their own components.

As to how OTAS will function, there will be essentially three parts to the system: prediction nodes, alert-issuing nodes, and clients. (1) The prediction nodes will take raw and/or processed seismic data from seismic stations that currently offer near real-time data over the Internet. Using tsunami prediction algorithms developed by others, or statistical and computational models developed by ourselves, these prediction nodes will determine the risk of whether (and where) a tsunami would occur resulting from a detectable seismic event. The prediction nodes would then pass the tsunami event information on to the alert nodes. (2) The alert nodes will be responsible for disseminating the tsunami alerts to the appropriate “customers” via various means. These nodes would keep track of some form of subscription information for each person who wished to receive an alert. Alerts could take the form of email messages, SMS messages, phone calls, HF and VHF marine radio alerts, etc., including our own OTAS clients. (3) The OTAS clients will be software running on our subscriber’s computers, web browsers, PDAs, Internet-able mobile phones, etc. that will receive alerts directly from our alert-issuing nodes. If there is some mode or method that someone wants to use to receive alerts, the open nature of OTAS will allow anyone to take our protocols and write a client to meet that request.

(2) Why are you involved? What is your motivation?

Two things brought me to this project. Before Robert Cringely wrote his call-to-arms for OTAS, I was a budding open-source developer looking for a project to work on, and I’ve been looking for a way to enhance my career as a scientist in my field. I’m currently working as a language editor for the Chinese Academy of Sciences, polishing the English for one of their publications. It’s a very good job, but at times I’ve felt the desire to do something more challenging. And while my job allows me to work on my own research, until now, nothing has piqued my interest. Working with other scientists and network application architects over the Internet on OTAS will give me a personal sense of satisfaction. Among all other software projects that I could choose to work on, this one in particular matches my skill set as a physical scientist. So that is exciting for me.

Now I had just finished reading the biography of Richard Stallman entitled Free as in Freedom. Stallman is the founder of the Gnu Project, which formed the Free Software movement that made the creation of the GNU/Linux operating system possible. In the book, Stallman talked about the early hacker ethic of “changing the world through software”. [Hacker here is used in the sense of someone who figures out how something works to improve it, especially computers, and not in the sense of a malicious computer user who tries to break things.] So the idea of changing the world through software really spoke to me. I had never thought of being able to use my computer skills that way until I read Stallman’s stories. Then, the OTAS project and the need for an open tsunami alert system appeared and I decided to sign myself in.

(3) What is your expertise? What sorts of things have you been up to in the past?

My expertise is in climate data analysis. For my Master’s research, I worked on building a 100-year dataset of weather data such as temperature, precipitation, wind, relative humidity, radiation, etc. for the province of Alberta in collaboration with Alberta Agriculture, Food and Rural Development. This dataset covers the entire province with data for every day in 1901–2000, making optimal use of the station data that was available. The dataset was designed for use in driving soil quality models, but is also useful to farmers, forestry management, real estate planning, utility use planning, etc. A beautiful full-colour climate atlas has since been published based on the dataset, to make the data accessible to many users.

For that project, I worked in the Statistical Climatology Group of the Department of Mathematical and Statistical Sciences of the University of Alberta.

The atlas is entitled Agroclimatic Atlas of Alberta 1971–2000, Agdex 071-1, by Shane Chetner and the Agroclimatic Atlas Working Group. (2003).

(4) How do you hope to help with this project?

I will be working on the tsunami detection and data analysis core of the system. I have no background in seismology, but I will apply what I know from my physics and statistics background to learn what is necessary for the project. We are fortunate to have Jose Simoes working with us, who is a seismology researcher from the University of Lisbon. At this point we are trying to determine what data are available to us on the Internet in near real-time. As you may know, some tsunamis are generated from seismic events and they propagate across the ocean at a predictable rate. So, the window between the seismic event and the “landfall” of the tsunami is the time period anyone wishing to make a tsunami alert has to work with. The sooner that data can be taken from the seismograph (and possibly tide-level) networks and placed into the OTAS system, the more time there is for people to respond to the alert. So the OTAS system will rely on timely raw or processed data that are publicly available. Once we determine what data sources there are, we need to evaluate their usefulness in tsunami prediction based on the way that OTAS is able to operate. The design and limitations of the OTAS system will depend, at one significant level, on the data that we can use.

(5) When did you join the project?

I contacted Robert Cringely soon after his “Wave of Change” article, and he put me in touch with Charlie Martin. The two of us have been corresponding as the infrastructure (website, mailing-list, etc.) was being built and new volunteers were signing up to help. The original article is dated December 30, 2004.

(6) Is there a political message behind this? The I, Cringley column sometimes says that this project is a substitute for a lack of political will. Is that how you see it?

Whenever a group of people come together voluntarily to build something for some purpose, you have a political movement. While not everyone on this project will care about the philosophical basis for the design and purpose of the system, we will be guided by the principle of openness and freedom. The system is being built by the people for the people. We will be sensitive to the needs of the subscribers to the OTAS alerts. If we don’t meet their needs in ways that are suited to them, we will have no users and our work will be futile. One of the first documents Charlie passed on to me was entitled “Why projects fail”, and it touched on that exact issue, among others.

An interesting facet of open source software development is the notion of a gift culture. No money is being exchanged and no leadership hierarchy is in place. Among ourselves, the developers of OTAS will be rewarded according to how much we give, by how much we contribute. And it is each individual’s reputation that will be on the line according to the work that he/she does. We are “competing for recognition” based not on what we control, but on what we give away. For a more elaborate essay on this topic, I point you to Eric Raymond’s book Homesteading the Noosphere, specifically the chapter, “The Hacker Milieu as Gift Culture”.

Ultimately, the reputations of the OTAS developers will be held out to the world based on the quality of service that OTAS provides. Failure to produce a working system will be one red mark. Missed alerts and false alerts are an entirely different beast. Plus we need to consider the question of liability since business and lives could potentially be disrupted by the normal and abnormal functioning of our system.

I just want to point you to this latest “I, Cringely” article in which Robert Cringely discusses why the Open Tsunami Alert System (OTAS) is an important development.

• Help Me Help You — Open Source Development as a Substitute for Political Will