Dan Bricklin's Web Site: www.bricklin.com
Home Writings	Trademarks and Domain Names Saying we'll just use normal trademark practices to solve the problems won't work. Here's why. < Open Access thoughts    Our photos as an edited record of our lives >
	This essay originally appeared November 1, 1999, in my log. I've been following some of the discussion about trademarks and domain names. To kill off "squatters", trademarks holders want to pull domain names away from people (often without much warning) and suddenly URLs will point to new places you don't expect. There is much emotion here. Stories of children with their domains named for their hobbies snatched by unfeeling lawyers. Pirates extorting large sums from technologically delayed companies. Etc. Bob Frankston, my old partner from the VisiCalc days as well as close friend and neighbor, has been working on an essay about "Safe Haven" names that give the benefit of today's domain names (i.e., not being tied to any specific IP address so you can change servers, and permanent so links won't break over time no matter what happens to legal names) without being tied to politically hot things like trademarks. Bob says: "Making handle assignment political threatens the basic 'glue' of the Internet." This got me thinking about trademarks and domain names, so I wrote this: Trademarks (marks of tradesmen) were used for many, many years to identify the source of items and to distinguish one tradesperson's crafts from another (like Revere's name or symbol on a pewter piece). The idea of trademarks was to avoid confusion on the part of the purchaser, and much of trademark law as I understand it rests on determining confusion between marks. Today the definition of a trademark says that it "...identifies and distinguishes the source of the goods and services of one party from those of others" (Basic Facts About Trademarks, US Patent and Trademark Office). The marks were originally symbols like boars, hog's heads, red labels, black labels, lions, etc., and eventually words and names. To avoid confusion, trademarks had to be unique and not confusing within their sphere, like the naming of drugs. Therefore, Cadillac cars were only made by General Motors, but there was no confusion with Cadillac dog food. Madonna the singer, Madonna the hospital, Madonna the religious icon -- no confusion. Smith the college, the person, the company. Bricklin the car, Bricklin the software, Bricklin the person. Each namespace (cars, dog food, singers, hospitals, people, etc.) had its own way to keep the names unique within its space. When you file for a trademark you must specify a Class of goods or services. (E.g., Class 7 is Machinery, Class 8 is Hand Tools.) You can't register a mark on "everything that ever exists". The problem with moving trademarks without thinking to the Internet is easy to see: convergence. We converge all namespaces that logically separate in the traditional world into one namespace in the domain name world. People used to policing their little space (cars, dog food, hand tools, whatever) now are all mixed up into one space. Entities not used to fighting for space (people, non-profits, informal organizations) are mixed in, too. Saying we'll just use normal trademark practices to solve the problems won't work. The checks and balances we had to keep each of the namespaces filled with unique names doesn't work when you throw in many existing namespaces with overlapping names and no established universal way of ensuring uniqueness or reassigning names. You can only ensure uniqueness when you start from scratch and keep it unique. We must leave separate namespaces with their own ways of avoiding name confusion. (Bob Frankston addresses these last two items through a new unique "handle" namespace and directories.) Adding to the problem is that the Trademark law concept of being "confusingly similar" is too subjective to use in the fast paced world of the Internet.
	Writings  Home  < Previous  Next > © Copyright 1999-2018 by Daniel Bricklin All Rights Reserved.