Hacker News new | ask | show | jobs
by 0xABADC0DA 5162 days ago
How so? Any jury, technical or not, should find 1A for Oracle.

"Has Oracle proven that Google has infringed the overall structure, sequence and organization of copyrighted works?"

As far as I can tell nobody disputes that Google lifted the Java API's structure, sequence, and organization from Sun, and the judge told them to consider SSO copyrightable. Most programmers think it's ok to rip off an API because otherwise the world will burn, or something, but 1A is asking what Google did not whether it was illegal.

My bet is the jury is held up on the fair use, either for implementation (1B) or documentation (2B). These are vague questions that can easily go either way. My bet is that they find 1B for Google (implementing the API is ok) but find 2B for Oracle (reprinting the API SSO is not ok). You have to implement the API to make it compatible, but you don't have to republish the API.

3 comments

I don't think either side disputes that Google implemented a portion of the Java API, which they copied from Apache Harmony. But they certainly seem to disagree on what qualifies as a meaningful definition of SSO, with Oracle arguing that both the entirety of Java and individual packages qualify as SSO.
It's not the jury's job to decide whether that qualifies as SSO, that's the judge's job. It's the jury's job to decide what people actually did, whether they did what they say they did, etc.
One of the jury's jobs is to decide whether Google infringed upon Oracle's protected property. It would seem like they'd require a clear definition of SSO to be able to do that.
What is "copyrighted works" in the question and can an API be copyrighted?
And do they still 'infringe' if it's classed as fair use?
I don't think the world would burn if APIs were copyrighted.

There are 'fair use' provisions and most companies would grant carte blanche to people to provide new implementations. I am actually struggling to think of an example other than Java.

So it would change nothing. In which case, why have them copyrightable at all? If companies aren't going to take advantage of it anyhow, it seems additionally pointless.

Fundamentally, the main reason to copyright something is to restrict others' use of it. Letting companies restrict their APIs' use would be unfortunate, so if it there is really no effect either way they should not be copyrightable.

SCO/Calderra would have loved to be able to claim ownership of libc for all *nix OSes. That's the kind of thing this case could be moving towards. Im sure somebody would love their tithes for SQL, too.

It could get very ugly.