On Wed, Nov 03, 1999 at 01:05:53PM +0100, Henning Makholm wrote: > What Raul seems to be getting at is that dpkg is presently the only > existing implementation og the command-line interface in question. > > His arguments apparently lead to a general principle: if, for some > protocol (a command line interface is a particular example of a > protocol, but one of the basic premises of the argument is that > such technical differences are irrelevant) there is only one existing > implementation of one side of the protocol, then every implementation > of the *other* side of the protocol automatically becomes deriviate > of the implementation of the first side. > > I don't think copyright law acknowledges such a principle.
Copyright law (by which I mean U.S. Copyright law) does indeed have such a principle, but it's formulated very differently: According to U.S. Copyright law, A ''computer program'' is a set of statements or instructions to be used directly or indirectly in a computer in order to bring about a certain result. Notice that this doesn't care about protocols or interfaces. If the instructions are used to produce the result they're part of the program -- this is a very inclusive statement. To balance this out, there's also a concept of fair use. Most uses of the command line interface count as fair use. I'm not going to go into this any deeper. I've posted that definition of a computer program something like a dozen times and most of the responses I've gotten don't even acknowledge the key issues. I worry that a second sentence would be too complicated for people. -- Raul