People tend to over-react. I’m not sure why; they must derive some enjoyment from it.
And I could definitely see that being the result of this: web2.0collage.com.
If pleasure from paranoia is your thing, that’s cool we can still be friends. But you should probably skip the rest of this post because I will point out that it cannot see which websites you’ve visited, so much as detect whether you’ve visited a website. A small, but in my opinion, very important distinction. The two would be equivalent if it checked every possible URL.
It works as follows:
There is a specific a tag. CSS defines its :visited style, so that when it points to a page you’ve been to it changes colour and size. Javascript then goes through a list, and for each entry points that tag to the corresponding URL. It then checks the style the browser has applied to it. If it matches the style defined for the CSS’s :visited, it decides you’ve been there.
If not? Maybe you hate that site and avoid it at all costs.
Nothing really special, just a clever-ish if ugly-ish (sorry) use of Javascript and CSS.
That’s a pretty cool idea, though? I can definitely see a market for it in advertising. E.g. say you start an advertising company that forwards small banners through your own constant server — as most of the banners out there are done. Whenever a banner is viewed you could passively build a web history of the visitor in the background, and store the progress in cookies, which would stick because your clients’ banners always happen off the same domain. Repeated viewing will allow the history to become very complete. Then you can use this to target advertisements, and all the information and gathering uses up the viewers’ cpu and disk rather than your servers.
most people would probably put a stop to that? I know I wouldn’t be cool with them javascripting up my CPU, especially not as inefficiently as that website does it 😛
I wouldn’t care about the privacy part though