The E-Data mentality strikes again

By Stephan Vladimir Bugaj
October 1998

Recently, a company called The Thinking Media was awarded a patent for client-side tracking of user interaction with banners. Only a Patent Office as thoroughly uninformed about both technology and the law could possibly grant a patent as broadly defined as "tracking client interaction with ads"!

Before coming here I worked at Bell Labs, and my boss tried to patent some of the ideas put forth in our two papers, including dynamic user interface elements that respond to the state of a program, graphic display of document age, my object model of Web and user profile data, and user interfaces that are purely iconic with no text representation. I have no idea how many of these patent requests ever made it out of the Bell Labs legal offices since our department was destroyed by internal politics, and I left Bell Labs and Lucent. While working at Bell Labs it was pretty scary watching people scramble to keep their ideas not only from competitors, but from each other, and to patent not only their ideas (good or bad) but mine. These requests were made even in cases where we never finished (or in come cases, never started!) an implementation. I was expecting them to ask me to leave my brain behind in a jar when I left the lab. Interestingly, our group asked the Bell Labs legal group to patent our system which was quite similar to what The Thinking Media patented. Our system was intended for use in tracking how long a user spends on a page, how much of that time is spent actively, and what the user is doing on the page. What we were working on was for entire Web sites, and theirs is just for banners, but the results are basically the same. Enforcement of such a patent would lock-out just about everyone else from any kind of statistics gathering enterprise online.

I find their description of what they're doing as "new measurement technology" particularly amusing. Their stuff is written in Java. Though I haven't bothered to decompile their code, I'd wager all they're doing is attaching a some kind of AWTEventListener based listener to each object (or working with processEvent and friends) and, in conjunction with a wall clock of their favorite flavor, tracking how many mouse movements, clicks, etc. are generated between the loading of the class (o if they're really accurate, they're starting their time from the creation of a graphics context for the main display frame) and a call to a finalizer which stops the clock (and in between, they have various events that trigger time markers that can be plotted, such as loading and unloading of new "pages" within the banner (add/remove components) and so on). Then, as they say below, it sends information to a central database. Whoa. Revolutionary! Big new breakthroughs there...

Whatever happened to patentable items requiring non-obviousness to practitioners in the given field? Even if one puts any faith in the US Patent system in theory, the types of patents that are being granted in practicality are - pardon the pun - patently absurd.

I just thought you'd like to know about this The Thinking Media patent... I know it caught my attention, it threatens the experimental work I'm doing into hypereconomics, as well as my practical livelihood of developing and managing technology for Internet advertising (though I must admit it's not my first choice of a livelihood, it pays the bills). What good is any kind of adaptive system on the Internet if it can not gather interaction statistics from the client in order to adapt without paying a royalty to The Thinking Media?