[Discuss] can you copyright an API?

[tmetro-blu at vl.com (Tom Metro)]

Edward Ned Harvey wrote:
>Tom Metro wrote:
>> Mostly is seems this is coming down to the issue of whether you can
>> patent an API...
> 
> Not patent an API.  Copyright the API.

Sorry, typo. Per the subject line and elsewhere, I obviously meant
copyright.


> Guess what, even the title of your book is copyrighted.

We have a copyright system that copyrights everything by default. But
there are limits...


> Even your distinctive and famous logo is trademarked and copyrighted. 

Trademarked, yes, and a specific rendering of the graphic design of the
logo is protected by copyright, yes, but the words themselves can't be
protected by copyright:

Can You Copyright A Single Word?
http://www.techdirt.com/articles/20110412/11105613870/can-you-copyright-single-word.shtml

That article happens to be about a case in Poland, but the same
limitation applies here.

http://en.wikipedia.org/wiki/Copyright#Obtaining_and_enforcing_copyright

  Typically, a work must meet minimal standards of originality in order
  to qualify for copyright...Different countries impose different tests,
  although generally the requirements are low; in the United Kingdom
  there has to be some 'skill, labour and judgment' that has gone into
  it. In Australia and the United Kingdom it has been held that a single
  word is insufficient to comprise a copyright work.

I'm not sure exactly where the threshold lies, and undoubtedly it is
something that has been established by court precedent rather than
stated in the law.


> These things can be used according to "fair use," but google would
> have a really hard time making a case that the API is "fair use," ...

Fair use is the exception that grants use of copyrighted works for
specific circumstances, but I expect Google will argue that copyright
doesn't apply to the API in the first place.

A commenter on the article I previously referenced offered this analogy:

  This is equivalent to writing a book called "2084" with the same
  chapter names as the classic "1984" however the actual text inside the
  chapters is different. So it looks like "1984" but once you actually
  read it, the store is completely different. And it's not even called
  "1984"!


> For one, they could have started with the GPL openjdk.

Yeah, I agree. Here are some quotes from a 2007 article explaining why
they chose to create Dalvik:

http://news.cnet.com/8301-13580_3-9815495-39.html

  "We wanted the platform to be open in a lot of different ways," said
  Mike Cleron, a Google senior staff engineer working on Android. "The
  idea is that anybody can come along and replace the pieces of the
  Android experience on a very fine-grained level. The existing APIs
  didn't really allow the level of openness we were hoping to achieve in
  Android."

They may have let the engineers rather than the lawyers make this
decision. If they had concluded that they wanted to throw out and
rewrite most of OpenJDK, and that they didn't want to be subject to the
politics of the Java Community Process, and had little expectation that
their changes would be accepted into OpenJDK, then doing their own
ground-up redesign makes sense.


> For two, why java?

Right, given how fond Google is of Python, they probably gave that some
consideration. But remember that even pre-smartphone, Java had a fairly
significant presence on mobile phones already.


> If you're going to the effort of writing the whole
> language from scratch, why not rename the classes and stuff...

Obviously they wanted to minimize impediments to developers coming from
other Java mobile platforms. Quoting from the same article as above:

  It should be noted that Google isn't working in a Java vacuum. For
  example, one of the OHA partners, Motorola, has helped lead
  development of Java for mobile devices, and Google wants to keep the
  Java programming experience familiar to developers.


> I think it's not so much the *volume* of alleged stolen code that matters,
> as the *value* of alleged stolen code.

If by value you mean the loss Oracle suffered as a result of the
infringement, yes. (As Jerry pointed out.)


> Why does a case like this go to court at all?  One reason, and one reason
> only.  The two parties couldn't negotiate agreeable terms for settlement.

Right. It has been reported in the news that settlement talks for this
case went on for a very long time.


> ...all google has to do is transliterate some words with other words
> in order to avoid future copyright infringement.

That would cover them for the API violation, though they'd still have to
pay damage for the past violation.

As far as the 9-line function that was found to be copied, they've
supposedly already rewritten it and released the new version in Android 4.0.


> Is it possible that Google...simply outright refused to pay more
> than $0?  Yeah, that's also possible, albeit less likely.  (Google
> might behave that way if they feel the legal precedent would risk 
> something of a larger stake..

It does seem plausible that they are fighting the case on principle and
to set a precedent on whether APIs can be protected by copyright.

 -Tom

-- 
Tom Metro
Venture Logic, Newton, MA, USA
"Enterprise solutions through open source."
Professional Profile: http://tmetro.venturelogic.com/