Re: About Computer Software and Patents

> >What is the argument that the software industry is different
> >from all other industries for patent purposes?
>
> The unique thing about software is that it simultaneously has the
> characteristics of a literary work, which would make it copyrightable,
> and of a useful object, which would make it patentable. There's also

Patent and copyright apply to different aspects of software. Patent applies to algorithms, whether they are embodied in software or not. Copyright applies to expression: the code, the screen designs.

> the mental process issue, that until the advent of software patents
> you couldn't get a patent on something you could in principle do in
> your head. There's a separate issue that the patent office doesn't
> have a clue about software prior art, but that's subordinate to the
> primary question.

The "mental steps" doctrine arose before software, and doesn't make sense applied to it. It was a creation of the PTO and conflicts flatly with the patent statute, which says that you may get a patent on an "art" which is what today we would call a "process." An algorithm is a process whether you can do it in your head or not; hence it falls within the subject matter of the patent laws.

That the PTO hasn't a clue about software is, I think, the fundamental problem.

> I've read Judge Keeton's decision in Lotus v. Paperback. He didn't
> understand the issues at all. The case hinged on the slash commands
> in 1-2-3. Paperback also copied the formula language, of course, but
> for some reason that wasn't an issue. Everyone agreed that the
> general L-shaped border around a spreadsheet was so general as not to
> be protectable, and the screen formats were different enough that
> they weren't an issue, either.

Protecting the menu structure takes "look and feel" to places it's never been before, places maybe it shouldn't go. Previously, L & F meant the way the screen looked in combination, perhaps, with the way you interacted with it. But at least there was always in previous cases some artistic expression visible on screen. I'm not sure I agree with Keeton on the menu thing myself, in other words, but I would certainly not say that he didn't understand the issues. He's an extremely bright and energetic judge.

> He reasoned that since the original selection of slash commands was
> arbitrary, it was a creative rather than functional choice and therefore
> was protected by copyright. Unfortunately, the slash commands define
> the macro language -- if you don't have the same slash commands, you
> can't run spreadsheets that contain macros, a functional problem that
> should be outside of the realm of copyright.

You can buy Lotus and run its macros. You can buy another spreadsheet and run ITS macros: there is no requirement that other spreadsheets be able to run Lotus macros in order for you to run macros of some kind. If you can't buy an illegal copy of a novel, that doesn't mean you can't read the novel; and reading is a function, isn't it?

> He rejected the functional
> argument, saying that it would be OK to have a conversion program that
> translated from the Lotus slash commands to a different non-infringing
> set. It seems to me that if you can do that, how about running the
> conversion program immediately after each character that the user
> types?

Good argument. I agree that the actual outcome of the case is troubling. I responded more to the "look & feel" criticism generally; it may be the case that this particular decision wasn't so hot, but that "L & F" generally isn't such a bad idea. I'm inclined to think that myself.

> Look-and-feel is not a suitable area for copyright, because it's
> basically a functional issue. All standard interfaces were once
> arbitrary, but aren't any more. Protecting the input langage to a
> program is like protecting the arrangement of levers and pedals in
> an automobile.

Again, if you are speaking specifically about the slash commands, that's a lot closer to "function" than some other things. But even so, copyright has for centuries protected works of function (as opposed to works of fiction or artistry). Until recently it protected phone books, and presumably still protects encyclopedias, lists, lamps, and whatnot. All these are "functional" works in some sense. The Apple II operating system was described as "functional" in Apple v. Franklin, but the court found that it was copyrighted nonetheless.

The notion that there is a "standard" interface now that is no longer arbitrary misses the mark. There are countless interfaces that are yet to be designed, and new ones are in fact being designed all the time. Magic Cap, e.g., is either a company or an interface or a program (I forget which) but it uses a new, non- desktop metaphor. It's coming onto the market fairly soon. Again, if consumers value an interface because it has acquired a kind of standards quality, then competitors are free to seek licenses to use the interface from the originator and capitalize on its value. It's done: Microsoft licensed certain parts of the Mac interface.

If a court wants to declare that a copyrightable interface has become so useful to the public that it can no longer receive copyright protection, then the court should ask the executive branch to initiate condemnation proceedings, set a fair market value on the interface (which presumably will be quite high, since everyone is asserting that they must have that interface and no other), and pay that price to the originator. That's the Constitutional way of turning private property over to public use.

            --Trotter Hardy

  +-------------------------------+------------------------------------+
  |  thardy[_at_]mail.wm.edu           |     Prof. I. Trotter Hardy         |
  |  Voice: (804) 221-3826        |     Marshall-Wythe School of Law   |
  |  Fax:   (804) 221-3261        |     College of William & Mary      |
  |  BBS:   (804) 221-1137        |     Williamsburg, VA 23187         |
  +-------------------------------+------------------------------------+