Douwe Osinga's Blog: July 2003

Thursday, July 31, 2003


A short update: I added Worldsizer to the projects. It's a flash applet that lets you resize the continents (or rather subcontinents) according to their size in wealth, people or military spending. The graphics are kinda crude, but it's cool nevertheless.

Tuesday, July 29, 2003

The four stadia of Internet Enlightenment

According to the Diamond Way Buddism, there are five paths on the way to full enlightenment. On the Internet there are four stadia.

  1. The novice, overwhelmed by the enormous amount of information and things available, becomes scared of losing the new found wealth and tries to hold on to it by printing it or saving it to a hard disk.

  2. After a while, this person reaches the first insight: the Internet doesn't go away. If you save the url, it is enough. If the information on the url is updated, there is a reason and you profit. If the url itself goes away, there is probably a reason for that too. This is the stage of bookmarking.

  3. The second insight is reached when the person realizes that urls are only temporary and that the best urls change all the time and that a good search engine is better then an outdated list of bookmarks. Want a site, just type in the best keywords.

  4. I'm not sure about the fourth stadium. I'm not there yet. Please let me know if you are.

Monday, July 28, 2003

A legal way to do Napster

In his latest column, Bob Cringely explains his idea how to create 'Son of Napster', a complicated scheme with shared ownership of CDs where everybody can make a copy of his own CDs of save keeping. Bob says it is legal, but it rather depends a lot on fair use, which I think would be hard to argue if you share ownership with a million other people of a CD.

A couple of months ago, I came up with my own scheme. It seems legal to me, but IANAL.

The central company, let's call it Napster II, is a storage and trading facility of CDs. If I have a CD, I can send it to the company and they will store it somewhere in a safe place and allow me online access to it. This way, I can listen to the CD anytime I want. I now, this is rather close to a similar scheme of MP3.Com, which didn't make it, but the difference here is that Napster II actually has the physical copy of the CD.

Every member of the scheme has let's say 5 CDs stored and can listen over the Internet to any song he owns. If a member wants to listen to a CD he doesn't own, he'll first has to buy it. This he can do by actually spending money, i.e. let Napster II buy it from a record store, or he can trade in one of his 5 CDs with somebody else. If I want to listen to the Beatles and I have a Stones CD, I only have to find somebody on Naspter II who has a Stones CD and wants to listen to the Beatles CD. We swap ownership (but the CDs remain where they are - safe with Napster II) and we both listen to the music we want. Napster II could charge a little something for the service.

Sunday, July 27, 2003

The power of the Internet

It seems if I go for a beer with a couple of friends, I'll meet more new ideas than if I would have watched TV for a week. But then, if I search the Internet for these ideas, most of them already have been implemented by some crazy guy and I'll end up with some more ideas just reading about the genial stuff other people have come up with. It seems that regular media has completely lost its capability of finding new things. TV broadcasts what people want to hear, what people already know. The Internet is full of new stuff. This might seem like a total trivial observation, but it spells trouble for humanity. We have the chance to explore new stuff from the convenience of our living room, but we'd rather watch reruns of Seinfeld.

Saturday, July 26, 2003

Killing an elephant

I came across this weird story. This year it is a hundred years ago that Topsy the elephant was electrocuted. Not by accident, but as a way of killing her because she had been bad and had killed three of her keepers. Apparently the last one had tried to feed her a lighter cigaret.

An interesting aspect of this strange tale is the involvement of the Edison Company. At the time, Thomas was trying to prove the superiority of his DC to AC, which was promoted by George Westingthouse. In the end he lost this battle, but one of his arguments was about savety. AC was supposed to be much more dangerous. Edison started to electrocute animals to prove how danerous it was. Financing and setting up the electrocution of Topsy was his last attempt to convince the American public.

He didn't succeed. We're using AC now most of the time and remember Edison for his light bulb and the gramophone and not for his cruelty against animals. Edison did convince the American public that killing by electrocution is an excellent way to execute convicts. It's still done in 10 states.


Friday, July 25, 2003

What's with Scandinavia?

If you're like me and you like statistics of all kinds, then you probably have noticed that when it comes to the well being of countries, it is usually the Scandinavians that win. Whether it is the richest country (Norway), or the least corrupt (Finland) or the most women friendly (Sweden) etc. They have been peaceful (after the stopped doing the Viking thing), pro-environment, pro-third world and generally nice to the world. They live long, are tall and handsome and speak good English. And of course they're awfully wired.

A month or two ago, there was a tape allegedly from Al Qaeda, calling the faithful to attack Norway. Nobody understood why. This in itself says something, usually you can come up with reason. But Norway is clean, except for the whale killing. Iceland is another example. Discovered by the Vikings, it was pretty much left alone for a thousand years. They were poor and cold and nobody paid much attention until suddenly their statisticians declared that Iceland is actually very rich. Fish and woollen sweaters did the trick. The sweaters are of course very expensive, but still.

The Scandinavian model is something that needs more study. Too bad about the weather, it would be perfect.

Saturday, July 19, 2003

Why America is losing the peace in Iraq

A comparision with the rebuilding of Japan and Germany is often made in regard to the conflict in Iraq. Why did that work so great and doesn't it seem to work now? First of all, it is still early. The Marshall plan only came into action 1948. The Nurnberg trials ended in 1949. Two month after the Second World War, nobody could believe it was really over. And the Americans are still in Germany.

But there is something else too. When the Alieds came to Germany and Japan, they came to conquer and punish. The Germans and Japansese had been bad. They lost the war and now everything was going to be different. That the alied forces took over the government was only logical. We'll take over and maybe, if you're nice, we'll let you be independent again. The collective guilt in the Axis made them accept the victors.

In Iraq the Americans have always said: we have nothing against the Iraqi people, it is Sadam Hussein we're after. The Iraqis are good, they are the victim of an evil regime. Then after the war, it is only logical that the Iraqi would say: if we're good and only Sadam was bad, can we please have our indepence right away, now that he's gone? There is no collective guilt.

Friday, July 18, 2003

Color scheme project

I added the color scheme thing as a project. Color evolution is alive and kicking. I also added the option from the color scheme page to see what a certain hue/saturation would do to the page. Try it, click the 'color this page' button below.

Thursday, July 17, 2003

Google holes

Steven Berlin writes an interesting piece in Slate about the holes in Google. He sees three:

  1. Shopping pages score above average

  2. Synonyms fill the top, driving out other meanings of a word

  3. Google drives us away from book knowledge

I think the first two are rather debatable, that is, it depends on what you want to find. Maybe a lot of people search the web to buy something and if one meaning of a word is that much more popular than an other, then that meaning should dominate.

The third point is a good one. A lot of humanities knowledge is still stored in the form of books and if we make Google our king of knowledge, we'll start ignoring this knowledge. However, another Internet powerhouse, Amazon, might rescue us. On the long run it would probably be worth it for them to scan and index the most popular books, thus allowing us to search the contents of books like we search web pages now. Of course you would have to buy the book to actually get to the content, otherwise this wouldn't make sense for Amazon (or they might show you the page they found the term on).

Wednesday, July 16, 2003

The gated community as the new Polis

According to the conclusion of the international conference on private urban governance, more then half of all new homes in the U.S. are guarded in some way. This might seem like a somewhat extreme reaction on the American anxiety, but the whole trend towards gated communities could actually be a good thing and a revival of the old Greek ideal of the city-state or polis.

The most obvious feature of a gated community is the fact that it is gated. You cannot enter and exit freely; you'll have to be a member or at least get permission. Gated communities are completely built on privately held land and can therefore make the rules. But here is the interesting thing, not only can they make rules about who can enter and who can't, but they can also make rules about a lot of aspects of social conduct: how to paint your house, when to put out the garbage and how to live together. Sounds scary?

Big government is receding everywhere. Parts of society are freed from too much interference and flower. Other parts are under pressure. In a society where the central government sticks to its core duties, gated communities might take over and over rules and structure for those who need that.

Thin government plus universal gated communities produces freedom. If you don't like the rules of the community, maybe you can change them by politics. But if there are gated communities everywhere, chances are that somebody already set one up with rules you like better, just like it is online. If you don't like a certain website, there's always one that is very similar, but that matches your way of thinking better. Gated communities are implementations of different ideas about the good life and therefore philosophical experiments. We should set up one based on the ideas of Kant.

What about the poor? Wouldn't they be ignored by the gated communities and slip through the cracks of the system? To some extend, probably. On the other hand, they would form a market for gated communities on their own. Our current society tends to throw the poor with the social unfit in the less desirable parts of town. A gated community for the poor would not have the golf courses the rich have, but it might offer a save place to live and a way for those who want to get out.

Tuesday, July 15, 2003

Googles political system

In a discussion on nettime, the question of the political structure of Google's algorithm came up. Before you reject the idea that there is any political structure to algorithms and especially to one as holy as Google's, think about it. Google self explains the page rank system in that they promote pages that have a lot of links to them from important pages in that linking to pages is like voting for a page. This is a political system.

The distinguishing feature of Google is that the role link popularity plays. This is independent of the actual search term (as far as we know), ie a page has a certain Google Rank and that helps the page, no matter what the user is searching for. This is not as much a populist structure, but more a technocratic. It is not the popularity among searchers that determines success, but the popularity among website builders/bloggers/corporations, what have you.

It made me wonder what other political system are possible in search engine land. The most successful one besides Google is the more capitalistic Goto (Overture) model, of course. Here the pages that are search just pay for their position in the ranking. The more you pay, the higher you end up in the listing. Note that this depends on the search terms used. This is obviously more a capitalistic political model.

The two models already seem to merge, with Google offering paid links (distinguishable from search results, but still) and Overture buying AltaVista giving it a more serious presence in the search engine space. This just in: Yahoo! bought Overture, so it probably is show time.

What else would be possible?

One could very well imagine a truly populist search engine, where not link popularity determines the position in the charts, but website traffic. Google couldn't really measure that, but ISPs like MSN or AOL could. Of course this would be a rather conservative search engine, making things popular that are already popular, but then, so is Google. Even parties like Kazaa, that are spreading already so much spyware, could include a little monitor to feed a search engine like this.

Another option would be a search engine where you can just buy general positions in a search engines, or the companies with the biggest market capitalization would score best in the search engines. After all, these companies are the most successful and should therefore be listened to.

A third option would be a Google variant where a searcher can rate websites. The Google rating following from links from websites I like is increased, the Google rating following from websites I dislike is decreased. This way I get results from websites that are liked by websites I like, etc. This would create a very fractioned search engine where everybody finds the answers she likes. I.e., if I don't like Microsoft, I will get the Microsoft sucks websites, if I like them I get

Drop me a note if you think of something else.

Sunday, July 13, 2003


I implemented a new color scheme for the site. All the colors are derived from one color using a mathematical scheme, loosely based on Goethes ideas about colors. Goethe and Newton had a deep dispute about the nature of color. Newton thought all colors could be broken up in a Red, Green and Blue component, while Goethe had a theory that was mostly esthetically oriented. Newtons theory was closer to the facts and is widely used in computerprograms.

Goethes ideas were not completely silly and are used in adapted versions in the setting up of colorschemes. Some people are naturals, other need a little help. I do this:

  • Choose the hue of a color from the color circle. Goethe had one, but the one in Photo shop will do just fine.

  • From this a very dark version is derived for the text and the borders (it is not exactly black) and two very light versions are derived for the background and the body of the texts (which is not totally white).

  • A much more intense version is derived from the opposite in the color circle for highlighting things

I'm not sure if it is a big improvement, so maybe I'm missing something. The plan is to evolve the colors of my website. The current color will be centrally stored. When a visitor arrives, he will be assigned a color that is close to the centrally stored one. When the user leaves my site, the number of visited pages is calculated and if it is high, the centrally stored color will be updated in the direction of the users color. Should be working anytime soon now.

it is working! From now on you'll the colors slowly changing.

Saturday, July 12, 2003


I just added a new project, GoogleBattle. GoogleBattle lets two search term battle by searching the top 10 of term1 OR term2 for occurences. Term1 OR Term2 returns all pages that contain either term. While on some search machines this would produces a list with at the top pages containing both, Google returns a list based on the page rank of the pages involved. By counting the occurences of the terms in these pages, you'll get an idea what the page rank of the terms is.

I also updated Archean. The Archean project explores self organization by multiplying the six dimensional strings in a matrix world with a transformation matrix, a little like Conway's game of life, but then in full color. I refined the way the colors are done a bit, making them less saturated and I added a 'breeding' mode. In breeding mode, four games of Archean run parallel, each with a slightly different underlying matrix. Clicking one of the games, replaces the other three with slightly different copies of the one clicked. This way, you can search for the nicest patterns.

Thursday, July 10, 2003

There can only be one

Like they say, there can only be one on the Internet. One big book store, one second hand store and one search engine to rule them all. According to my brother, now we have an absolute measure to decide who is the one. Who is it. Bob, you mean the Bob, just means the Bob with the number one position on Google, the I feel lucky Bob. It's an interesting thought. For ages man was the measure of all things. Now Google is. Google has become a mediator in disputes. Who is the most famous painter? Picasso. Why? He has more hits on Google then Van Gohg or Rembrandt. Who is the search engine? Google (though a month ago it was still Alta Vista). The bank? The World Bank.

So who is the the, or what is the word on Google with the most hits and who has the number one spot on that? The word with the most hits, is 'the', fittingly enough. The Onion holds the number one spot for 'the', just before the Whitehouse.


Tuesday, July 8, 2003

We don't need no education

Pink Floyd sang it so many years ago. Secundary education, do we need it? If you ask around, you notice nobody remembers a thing what they were taught back then. Some English maybe, but it could be I learned that from movies and listening to Bruce Springsteen anyway. Chemistry? Forget it, people talk about 'natural ingredients' as if Friedrich W? never lived. Mathematics? People have nightmares about it, that's all.

The sad fact is, you can't really teach people something they don't want to learn and at the age people are when they suffer secundary education, they're usually not very interested in learning. Furthermore, the stuff on offer is not very intereseting nor is it very useful. General knowledge is all very well, but it is not useful, outside the realm of trying to impress other bloggers (although they would assume I got the Fredrich W? thing from Google). In reality high school doesn't really prepare you for anything. It keeps you busy (and not even that in some cases) until you're old enough for university.

Secundary education is a faulty solution to a problem we can't solve. You have this four to six year period in which you have to do something with the kids and you really want to prepare them for the future by teaching them all kinds of things. It sounds all very much all right and so we ignore the fact that most of the time (teachers & pupils) is just wasted. The studied materials are forgotten and mostly irrelavant anyway.

So what can we do? Learning languages is something that might be useful. In all honesty, I did learn English and German on highschool and am reasonably fluent in either (also learned French though). If we would sent all kids two times for half a year to a foreign language country, that would help. Also, I would have liked to learn the stuff that took me ages, like how to sell stuff (yourself for example), talk to girls and other social skills. That is something a 16 year old can use straightaway.

Monday, July 7, 2003

Language evolution and accents

A couple of weeks ago, I wrote about language evolution and where complexities in languages come from if they seem to go away when times goes by. This is high level, grammatical evolution, taking place over hundreds if not thousands of years ago.

Today it struck me you could study the relationship between languages in different way. The speech accent archive of the George Masson University has a large collection of non english speakers, pronouncing the same paragraph of english texts. The difference is pronciation is probably related to their own language. The people from the archive already mapped the texts on ipa. It would be interesting to write a little program that would find the shortest paths between these fragments and try to order the native speakers based on how their pronunciation differs. Would the resulting map of languages resemble the normal language maps?

Sunday, July 6, 2003

If index followers win, who will lead?

You can't beat the index when it comes to the stock market, that is, not everybody can, because then the index would be doing better, the index being the average. People investing in active management funds are therefore suckers, unless they somehow know which active funds are going to perform above average (and on average they don't). According to morningstar, an active fund manager takes 1.45% of the capital per year. And most people in active funds, try to beat the market by getting into the best active funds, which means switching, which costs even more money. On average, tracking the index is the best for the individual investor.

Right now, most people don't seem to realize. 90% of the mutual fund money goes into actively managed funds, only 10% follows the index. But what if people would come to their senses? What if all people would put their money in index-trackers? Who would then lead the indexes, i.e. how would the index know where to go to?

For that matter, how does it work now? In a perfect market, all available information is already incorporated in the market and the index always knows best. But before the information goes into the market, somebody has the opportunity to profit. Insider trading, you can call this. It happens when somebody knows something, before the market does, but it also happens when some people just have faster access to trading platforms when news breaks.

The lesson: follow the index at all times, unless you have insider information. But if that holds, in the long run only the people with insider information will trade actively, making it very easy to bust them.

Saturday, July 5, 2003

The coming Internet ice age

Slowly but unstoppably, the Internet is freezing over. Sites that used to be changing daily with fresh content produced by professional writers broadcast the same old news anno mid 2000 over and over, with their owners only sometimes posting a little article written in their free time. It is unavoidable. The bubble pumped so much money in the Internet that the whole info-ecosystem was growing like a rainforest on steroids, like there was no tomorrow. There was.

Between the dead and frozen trees of the old days, a new and more nimble system is developing. Blogs connected through the long and thin threads of RSS have developed their own maze of complexity. Paid for content sites are growing again. Amateur communities that never where touch by the bubble frenzy, live like they did before the boom. But the great dying hasn't finished yet. The old content empires aren't all quite deserted. Some are overgrown by a myriad of pop-up, pop-under and other in-your-face advertising. It will take a while before the ice age in its full strength is up on us.

Many searches for practical information still bring you to sites from 1999-2001. Two to five years old, which is okay for a lot of purposes, but what will it look like in ten years? Will the world wide web be seen as a museum of information of the end of the twentieth century? Too many old websites will devalue the whole medium.

Then there is Google. It seems a power for the good, crawling the web searching for relevant content. But it won't save us from the coming Ice age, not the way Google works right now. Google sorts the web by linking. The more a page is linked to, the better it scores. But the old, dead websites are linked to a lot and they link to each other. Not only keeps Google directing traffic to these frozen dinosaurs, in subtle ways it helps the over icing go on.

Bloggers like to spice up their websites with links to relevant terms, but they are often too lazy to really research something, or rather they are writing about the ice age and not about the relevant term, so they don't have the time to research this term. So they look it up on Google and paste the link to the first hit, keeping a dead ecosystem popular.

Is this all bad? Maybe not. In a way it will help the new forms of information exchange and teach us that pure Internet forms just work better.

Friday, July 4, 2003

A Pan European Language

The issue which language(s) to use in the European Union has taken up a lot of time already and will probably do so more in the future. This, of course, is silly. Europe already has a language that everybody speaks. Everybody except the British, but they haven't joined the Euro and they still might. These things take longer with the British, so they might learn to speak Euro-English yet.

Put a Frenchman and a Czech in a room and they will have no trouble at all understanding each other. Add somebody from Liverpool and they won't understand a word of the new guy. The accent is wrong, the words are akward, people from Great Brittian have great difficulty making themselves understood in Europe.

It sounds over the top, but it is true. And the EU should draw its conclusions. Euro English should be the European language, and should be taught in school instead of the strange mishmash of American and British we have now. Some money should be spend on research into this new language and more importantly, we should come up with European way of spelling the language. Euro English may be derived from British, there is really now reason to take the spelling, which is as arcane as the pound and mile.

Thursday, July 3, 2003

Battling spam

It seems that everybody and his brother has solution to the growing spam problem these days. At out office, three out of every more mails is spam. We filter and delete by hand, but nothing is perfect.

The solution I'm proposing won't work for a small outfit like ours, but should work for operations like hotmail, with huge mailservers. You set up a couple of bogus e-mail accounts that you have spam bots harvest, respond to dictionary attacks etcetera. Make sure with these accounts you're on every spam list out there.

Then when a mail comes in, you check whether this mail has already been received by one of the bogus accounts. If so, it is spam. If not, probably not. You might have to hold the mail for a little bit, in order to give the bogus accounts some time to catch up.

Wednesday, July 2, 2003

Driving and riding

It seems that the days of the great trains are nearing the end. The car and the plane are eating into the realm of the iron horse. The door-to-door convenience of the car wins for the shorter distances and for the longer distances the plane is much faster and usually cheaper. Lately planes have become a lot cheaper, so that they beat the train even at distances as short as a few hundred kilometers. EasyJet and Ryan Air fly for under a hundred euro to lots of places and not only if you book weeks ahead.

Although it would make me sad to see the train go, I do believe we should start thinking now what to do with the legacy. We could use the land for other purposes and indeed it might very well be the case that already the ground on which tracks are build are worth more then the whole railroad system, but there is something else to it, that could be worth a lot more: the right of way. The most valuable part of a railwaysystem is that it connects the centres of major cities. Cut up the network and you lose the value.

So what will we do with this network after the train has gone? One option might be automated cars. Technically, a lot is already possible. The main problem is: how to get critical mass. If everybody would have an automated car, then it would be worth it to build automated highways, ie highways that let cars know where they are, where the cars are etc. This would reduce traffic accidents greatly, increase speed and if some sort of grouping could be done, reduce environmental stress too. But there are almost no automated cars, so why build automated highways and since there are no automated highways, why build automated cars.

Railroads could be an inbetween. In a way, railroads are already automated highways, only for trains. The railroads know pretty much what is going on. If we could build cars that could also drive on railroads, but then in automated mode, communicating with the automated systems of the railroads already in place, we would have something. We could start with long distances and at night. You drive to a railwaystation, put your car in automatic, go to sleep and wake up the next day, whereever you want. The same as with car trains, but without the trains and with a point-to-point option. From there the system could be expanded to a really general system. Finally when the trains have disappeared we're left with something better.

Tuesday, July 1, 2003

How to make money counterfeiting very hard

I came across this article that describes a scheme to stop money counterfeiting and then it hit me that this can be done rather simpler. Here it goes.

The central bank should keep like four times the amount of money in their vaults and they should keep track of the serial number of these notes out there and in the vaults. Serial numbers should be chosen random from a large range in such a way that the chance a random number would denote an actual number would give you a one in million chance of getting it right. The bank should also allow to circle the notes as fast as possible from the vaults into society and vice versa. Furthermore, a simple and quick (but not to quick, say a 1 sec wait) digital service would allow people to check whether a certain number is a currently in circulation (i.e. exists and is not in the vaults).

Now if somebody would want to counterfeit money and spend it, he would first get a valid serial number. The only practical way to do this, would be to copy one from an existing note. If the counterfeiter would succeed in getting enough numbers to make it worth while, spending the false notes would still give him a chance of 80% of getting caught, because the serial number would probably no longer be in society.