API Copyrightability in Oracle America [sic] v Google

Judge Alsup announced his findings on copyright as an order in the Oracle vs Google case. I’m surprised that I haven’s seen any discussion in my usual social channels. I can’t imagine why not.

The narrow conclusion is that the Java API is not protectable by copyright.

I found the judge’s order to be extraordinarily clear and free of legal jargon. I felt that he very much intended it to be read and understood by people in my field. You can read it yourself at Grocklaw

The principle seems to be that useful activity comes under the domain of patent law rather than copyright law.

  • Copyright asserts a monopoly for 95 years without any decision or action or finding by the government, but only applies to non-essential creative activity, and only to a single fixed expression of that creativity. It never covers the ideas behind that expression, methods of operation, or names.
  • Patent grants a monopoly for only 20 years, and only after an examination by a government agent, but it can cover useful methods of operation, beyond the specific embodiment being commercially protected.

Two silly examples clarify the distinction.

  • If an application UI layout is an efficient way to accomplish the task, then it is patentable (if original and non-obvious), but not protectable by copyright. However, if the UI makes the user jump through unnecessary hoops to get anywhere, than this creative diversion is copyrightable. If not patented, a competitor is free to copy the efficient design, but slavishly copying the silly one could be a copyright violation.
  • A security protocol that operates simply and directly may be patentable, but not copyrightable. If not patented, a competitor can reverse engineer and duplicate the protocol without violating copyright (as long as they don’t literally copy the high-level language source in which the copyrighted program was implemented). However, if a copyrighted program intersperses spurious nonsense that is ignored by the protocol, then a reverse-engineered duplicate that slavishly copies the creative nonsense would be a violation.

Being an engineer rather than a lawyer, it seems to me that these examples show the law to be based on the wrong principle with which to benefit society, even if they do provide the right answer in the more common cases to date. In my opinion, “efficiency” is a weak and manipulatable yardstick. For example, if inefficient nonsense achieves monopoly business protection for 95 years without legal costs or examination, then it has indeed achieved its “useful” purpose (for the claimant, if not for the user).

But I do not feel this silliness makes Judge Alsup wrong. I think that he arrived at exactly the right answer in the case before him, and I think he does us a great service in how he laid out his reasoning. I do worry that these examples will lead to more software patent filings rather than less, and that this will be a bad thing. I worry that users will suffer through deliberately inefficient designs, just as they already suffer through impenetrable and overly comprehensive usage agreements. I worry that smaller, more independent and more lithe developers will be buried by large corporations that cannot compete on innovation. However, in this case that’s all a consequence of how the laws are written, rather than how the judiciary works with what they’re given.

Interesting (to me) that in the whole order, there is no mention of open source or any notion of “free” licensing of any kind.

This entry was posted in Inventing the Future, Philosophical business mumbo-jumbo. Bookmark the permalink. Both comments and trackbacks are currently closed.
  • Connect With Us

    Follow Wetmachine on Twitter!


If you do not have an account: Register