LWN: Comments on "Google guilty of infringement in Oracle trial; future legal headaches loom (ars technica)"

Google guilty of infringement in Oracle trial; future legal headaches loom(ars technica)

nye — Mon, 14 May 2012 10:27:36 +0000

>You seem to be upset that the question was "unfair". Of course it was: facts aren't fair, they're just facts. No one sane would have found otherwise. But the Jury still had to issue a finding.

I think you have me confused for somebody else. I'm not upset at all; I'm just trying to understand the point of this - I wasn't previously aware that facts which are not in dispute still had to be decided by a jury (in this country we wouldn't even have a jury in this case).

Google guilty of infringement in Oracle trial; future legal headaches loom(ars technica)

ajross — Thu, 10 May 2012 15:31:05 +0000

> What was the point of asking the question in the first place if everyone already agrees on the answer?

Because it's a court case, not a debate, and courts have rules. Among them is the requirement that issues of fact (like whether or not infringement occurred) must be decided by a Jury.

You seem to be upset that the question was "unfair". Of course it was: facts aren't fair, they're just facts. No one sane would have found otherwise. But the Jury still had to issue a finding.

Google guilty of infringement in Oracle trial; future legal headaches loom(ars technica)

mikov — Thu, 10 May 2012 14:58:28 +0000

Thank you! That was exactly what I was wondering.

Google guilty of infringement in Oracle trial; future legal headaches loom(ars technica)

nye — Thu, 10 May 2012 14:54:30 +0000

>You're confused about how this works. This isn't a criminal trial. The jury doesn't decide the case, they decide the facts. The question of whether APIs are copyrightable is not a "fact", it's an interpretation, and the judge will decide that.

Given the question of whether APIs are copyrightable is not the issue, and given that the question of whether Google's usage falls within fair use is also separate, it seems that the only question the jury were asked to consider is 'did any Java code (where 'code' includes headers) go into Dalvik?'

What I'm not seeing is any way the jury could have decided 'no'. This doesn't even seem to be contested; of course the header definitions were copied. What was the point of asking the question in the first place if everyone already agrees on the answer?

API copyright nightmare

NRArnot — Thu, 10 May 2012 12:21:00 +0000

If they are copyright-able, how far back do we go? Presumably to the dawn of the digital computer, because copyrights last for many years after the death of the author.

Someone must have been the first to implement a stack for procedure calls in assembly language. That's an ABI. If it's copyright-able, every piece of code written since then that uses a stack to pass arguments and save return addresses is in violation!

So say goodbye to everything containing a CPU.

Hopefully the judge will realize that the ultimate conclusion of allowing ABIs to be copyrighted is much the same as allowing a toddler to play with a fully-armed H-bomb.

Ars got it wrong - go to Groklaw

charlieb — Wed, 09 May 2012 13:17:34 +0000

> I'm comparing this to the SCO case (series of cases), which went on
> for what, 7 years?

You speak about that case in the past tense - but it is still ongoing.

Google guilty of infringement in Oracle trial; future legal headaches loom(ars technica)

andreasb — Wed, 09 May 2012 01:11:56 +0000

> That's like telling a Jury to decide: Assuming that smiling is copyrightable and Joe smiled first did Tom then violate copyright in this photograph. Safe to say if the Tom is seen smiling then yes he did violate copyright.

Now this is off-topic and nitpicking, but still… Your example would work if it were about patents. Unless I'm mistaken, copyright is about the actual copying of works, not their similarity. If Tom has never seen Joe's smile he doesn't violate copyright with his own smile since it is not based on Joe's work — even if both smiles looked identical. In fact he would have his own copyright on that smile.

ISTR there being an actual case of two independent copyrights on mostly identical musical works.

Ars got it wrong - go to Groklaw

cortana — Tue, 08 May 2012 22:32:59 +0000

I'm amazed (and pleased) that things are moving this quickly. I'm comparing this to the SCO case (series of cases), which went on for what, 7 years?

Ars got it wrong - go to Groklaw

dbruce — Tue, 08 May 2012 22:13:28 +0000

I know there are comments above which basically say the same thing, but later posters don't seem to have seen them.

This was a nearly-complete win for Google. To paraphrase, the judge instructed the jury that "If we assume for a moment that APIs are copyrightable (which I will decide later), does that mean Google infringed anything"? The jury responded that "If APIs are copyrightable, then yes, Google infringed them. Plus, Google shouldn't have copied that 9 line function that slipped into Android by mistake and has since been removed".

So, Oracle is dead in the water as far as copyrights are concerned unless the judge accepts their radical idea that APIs can be copyrighted, which would fly in the face of how programming has worked for decades. This judge clearly "gets it", and is well aware of the recent EU decision on this very subject. I think Google will be completely off the hook except for the nine lines of now-removed code of the 15M LOC in Android. Perhaps the judge will award Oracle 0.0000006% of the price of a few hundred million *free* downloads of the Android system.

Google guilty of infringement in Oracle trial; future legal headaches loom(ars technica)

smoogen — Tue, 08 May 2012 19:35:40 +0000

In some cases it is. To form an informed opinion on this, one needs to know if Federal law, the 9th Circuit or the specific district rules asks for or requires a judge to put forth such questions. Without knowing that your original question is not informed opinion but what we would want the world to be like.

Google guilty of infringement in Oracle trial; future legal headaches loom(ars technica)

wahern — Tue, 08 May 2012 18:54:54 +0000

I meant the _right_ to a jury trial is required. The right is waivable at the discretion of the defendant.

Google guilty of infringement in Oracle trial; future legal headaches loom(ars technica)

wahern — Tue, 08 May 2012 18:53:15 +0000

A jury is required in this case because of the 7th Amendment to the United States Constitution:

http://en.wikipedia.org/wiki/Seventh_Amendment_to_the_Uni...

This makes the United States the last country, I believe, to keep juries for civil cases, and extremely difficult to change. Most of the individual states still have juries for civil cases, too, even though this is usually not required by the state's constitution.

Google guilty of infringement in Oracle trial; future legal headaches loom(ars technica)

wahern — Tue, 08 May 2012 18:47:12 +0000

The jury doesn't decide what constitutes Fair Use. The judge describes the legal elements of Fair Use, and how they apply in the case. It's the jury's task to decide whether the facts fit the elements, including whether they believe the facts as presented by Google.

In the US, juries decide facts, judges interpret the law

wahern — Tue, 08 May 2012 18:43:09 +0000

A good book which analyses the contemporary and historical use of juries in the context of an overall critique of the contemporary criminal justice system is The Collapse of American Criminal Justice by the late William Stuntz.

There are undoubtedly better books describing the history of the jury system, but according to the thesis of the book various changes in the the criminal system in the United States has turned juries into rubber stamps. Jury conviction rates are significantly higher (double or more) today than a hundred years ago. Juries used to have far more discretion than they do today. Now the laws are so detailed, and their definitions so all-encompassing, that juries are given very little leeway to show leniency or to provide the defendant with any true benefit of a doubt.

Google guilty of infringement in Oracle trial; future legal headaches loom(ars technica)

b7j0c — Tue, 08 May 2012 18:27:34 +0000

congrats RMS, you were right....the "java trap" is real

Google guilty of infringement in Oracle trial; future legal headaches loom(ars technica)

SEMW — Tue, 08 May 2012 17:59:59 +0000

It does seem slightly odd. In England juries in civil trials have been abolished except in cases where you want juries as guards against an authoritarian criminal justice system (malicious prosecution or false imprisonment), fraud, or defamation: http://www.legislation.gov.uk/ukpga/1981/54/section/69 .

In the US, juries decide facts, judges interpret the law

job — Tue, 08 May 2012 17:30:25 +0000

I had no idea! (I seem to remember vaguely from school that the US system used Montesquieu's principles but this seems to run counter to those.) The possibilities sounds endless, but I guess it isn't practiced much?

Google guilty of infringement in Oracle trial; future legal headaches loom(ars technica)

iabervon — Tue, 08 May 2012 16:52:19 +0000

There's a potentially-significant difference of attitude in whether he said "assume that this is copyrightable" or "ignore the issue that this may not be copyrightable" or "don't consider whether this is copyrightable or not". I would guess that any judge would say the third of these to a jury rather than either of the others (least prejudicial), while a paralegal would report it as the first (most logical). But I haven't read transcripts so I don't actually know.

In the US, juries decide facts, judges interpret the law

clugstj — Tue, 08 May 2012 16:24:43 +0000

If all you know about the jury system is what you've read in the news, then I'm not surprised that you find it strange. All news outlets are biased - at least toward what they think you will be interested in hearing about. The news is certainly not how you should learn about how some process normally works.

Google guilty of infringement in Oracle trial; future legal headaches loom(ars technica)

tseaver — Tue, 08 May 2012 15:47:37 +0000

The really weird thing here is that the question of fair use is deemed a
"matter of fact", to be settled by a jury, while the question of whether
an API is copyrightable is a "matter of law", to be settled by the judge:
they both require similar depth of understanding of copyright law.

Google guilty of infringement in Oracle trial; future legal headaches loom(ars technica)

amacater — Tue, 08 May 2012 15:35:14 +0000

In fact, if you read carefully, this is _exactly_ explained by Groklaw.

Infringement is a matter of fact - so the judge is saying to the jury:
"assume for the moment that Oracle have copyright: if so, on the facts
in front of you, is it a fact that Google infringed? If they did infringe, did they have any defence"

That's a finding of fact by the jury: if Google never infringed, Oracle have no case. If google have a valid defence, Oracle have no case.

This gets the finding of fact, which is all that the jury can do. _When_ this case goes to appeal, they can leave aside the factual questions put to the jury - "Jury nullification" - and concentrate on the issues of law.

Then the judge goes on to establish the law as to whether APIs are actually copyrightable.

Then the judge will go on to decide the patent issue

In the US, juries decide facts, judges interpret the law

drag — Tue, 08 May 2012 14:53:24 +0000

> I must say the jury system is very strange to an outsider. Lots of blatantly apparent problems: your chances vary wildly depending on whether you are famous, from a racial minority etc. If the idea is to protect against dysfunctional judges, I'm pretty sure there are better ways.

The only major advantage of the Jury system in the USA is that the Jury has the legal ability to nullify laws. If a Jury decides that a law or a ruling is unjust they could declare the defendant not guilty, even if he is clearly guilty of breaking the law.

Google guilty of infringement in Oracle trial; future legal headaches loom(ars technica)

jpnp — Tue, 08 May 2012 13:16:20 +0000

I guess that even Judge Alsup knew creating precedent on the copyright-ability of APIs has major implications and was trying to avoid having to do so. If the jury had cleared Google, he would have succeeded; as it is he'll have to make a decision.

In the US, juries decide facts, judges interpret the law

Wol — Tue, 08 May 2012 12:14:21 +0000

Note that the original idea of "a jury of your peers" has already been captured ...

The US habit of throwing people off pretty much *ensures* it is NOT a jury of your peers, as originally constituted.

Cheers,
Wol

In the US, juries decide facts, judges interpret the law

sorpigal — Tue, 08 May 2012 11:37:57 +0000

I'm thinking also that the judge was hoping the jury might come back with a "No" for the infringement question, which would mean he wouldn't have to rule on whether APIs are copyrightable because the outcome would be moot. For sticky issues like this judges would probably just as soon leave the question unanswered, if they can, in hopes that the industry works it out peacefully without ever having it come to a ruling the outcome of which can only be bad for one group or another.

In the US, juries decide facts, judges interpret the law

nix — Tue, 08 May 2012 09:06:25 +0000

If the idea is to protect against dysfunctional judges, I'm pretty sure there are better ways.

That's one of the purposes. Another purpose is to ensure that legislators and judges cannot drift too far away from the average man -- and if anything the US habit of throwing off juries anyone who has detectable skills helps here, odious though it is. It's as if they're de-eliting the jury pool. (Not that the US system of jury selection doesn't have other huge problems -- jurisdiction shopping, groundless but nonetheless useful appeals to local pride...)

I also note that people have been looking for a better system than juries (in an ad-hoc intermittent fashion) for most of a millennium, and nothing obviously better has emerged. Everything else anyone has tried is prone to capture by one or another interest group or power bloc.

In the US, juries decide facts, judges interpret the law

job — Tue, 08 May 2012 08:17:50 +0000

I must say the jury system is very strange to an outsider. Lots of blatantly apparent problems: your chances vary wildly depending on whether you are famous, from a racial minority etc. If the idea is to protect against dysfunctional judges, I'm pretty sure there are better ways.

Anyways, I've learned not to put to much weight in these kind of rulings. I'm sure the gaping holes in the ruling is reasonable for someone who understands the legal system, and that it will all be explained in time. There is little alternative to Google winning this in the wider economical perspective.

Google guilty of infringement in Oracle trial; future legal headaches loom(ars technica)

job — Tue, 08 May 2012 08:10:28 +0000

Oh! That would be fun! I bet IBM is polishing their lawyers as we speak...

In the US, juries decide facts, judges interpret the law

farnz — Tue, 08 May 2012 07:56:32 +0000

Note also that it's clear that Judge Alsup is conducting this trial with a view to a possible appeal.

By getting the jury to determine facts that may become irrelevant depending on how he decides the law, he's set up to avoid having to reconvene a jury trial with a new jury in the event that the appeals court tells him he made a mistake in his interpretation of the law.

From the outside, it looks like Judge Alsup is very aware of how important any precedent set as a result of this case will be, and doesn't want questions of fact distracting the appeals courts from the questions of law.

As a nice side effect (and I think he's considered this, too) it ensures that when the appeals court is asked to rule on matters of law, it's not completely abstract; the appeals court will be able to look at their decision in the light of a real jury's findings, and will be aware of the full implications of their decision - there won't be any "well, this is how the world should be, because I expect a jury would obviously see this case as fair use" in the appeals world.

Google guilty of infringement in Oracle trial; future legal headaches loom(ars technica)

Seegras — Tue, 08 May 2012 07:12:25 +0000

You're being conservative.

If APIs are copyrightable, all the Hell will break loose.

C - C++ - Java... EVERYONE will be in a mess. SQL - PL/SQL ... especially Oracle ;)))

In the US, juries decide facts, judges interpret the law

cmccabe — Tue, 08 May 2012 05:16:16 +0000

This could almost be a good thing, IF we get a strong precedent that APIs can't be copyrighted. That's a pretty big if, but I guess you have to take your comfort where you can find it.

Google guilty of infringement in Oracle trial; future legal headaches loom(ars technica)

jjs — Tue, 08 May 2012 02:30:22 +0000

Actually, like all good lawyers, Google had multiple defenses:

1. The APIs aren't copyrightable
2. Even if they are, we didn't copy enough to matter
3. Even if we did, we did it under fair use
4. Even if we didn't meet fair use, Sun had already given us permission via CEO and other statements

Probably others as well.

In the US, juries decide facts, judges interpret the law

JoeBuck — Tue, 08 May 2012 01:04:20 +0000

IANAL, but I'm paraphrasing what I was told by a judge when I was on a jury pool: juries decide what the facts of a case are, judges lay out the rules about what they are to decide (for example, a judge will explain what premeditation is, and the jury has to decide whether a killing was premeditated).

As I understand it, the jury was asked: assuming that APIs are copyrightable, did Google engage in copying? The jury, of course, replied "yes": we all know that they had to do this to implement Java. They were asked a second question: given some rules that the judge explained to the jury about fair use, was the Google copying (if any) fair use? In this case, the jury answered "we can't decide" (they couldn't agree).

So it's hard to say based on this that Google lost. Had the jury clearly said that the copying wasn't fair use, they'd be in trouble.

Google guilty of infringement in Oracle trial; future legal headaches loom(ars technica)

ajross — Mon, 07 May 2012 23:59:06 +0000

You're confused about how this works. This isn't a criminal trial. The jury doesn't decide the case, they decide the facts. The question of whether APIs are copyrightable is not a "fact", it's an interpretation, and the judge will decide that.

Think of this as an optimization: getting a finding of facts out of the jury now (i.e. "If I later find that the API is copyrightable, did they infringe?") allows the jury to be dismissed earlier and gets the media out of the courtroom.

Google guilty of infringement in Oracle trial; future legal headaches loom(ars technica)

blitzkrieg3 — Mon, 07 May 2012 23:51:19 +0000

From your link:

> [Update 6: The judge has stated, pending judgment as a matter of law, that there is "zero finding of copyright liability" other than the 9 lines of code to which Oracle's damages report attributes no value. A good day for Google overall.]

Google guilty of infringement in Oracle trial; future legal headaches loom(ars technica)

drag — Mon, 07 May 2012 22:51:17 +0000

> Don't you agree that "assuming" that APIs are copyrigt-able, when Google's major defense is that they are not, further prejudices the jury against Google?

Not really. Not unless the jury was made up of morons.

The Jury's job was to decide whether or not Google violated copyright, assuming that APIs are copyright-able.

That's like telling a Jury to decide: Assuming that smiling is copyrightable and Joe smiled first did Tom then violate copyright in this photograph. Safe to say if the Tom is seen smiling then yes he did violate copyright. If the guy was not smiling then it wouldn't make the Jury biased against the photograph, would it?

It doesn't seem like much of a surprise at all. The only surprise was that the Jury was so mixed over it.

It seems very true that Android does use some Java APIs. That was one of the major points behind Android in the first place... Java is very popular in embedded space and thus a lot of developers were already familiar with developing for Android before Android ever came out. In addition Android works by using a Sun Java compiler to create bytecode that is then transformed to a compatible format for Dalvik.

So Google's defense in this case was that they added so much to the API that it was no longer derivative.. that their version constituted fair use. Which seems like a weak argument.

Now the Judge has to decide whether or not APIs are copyrightable.

The Judge was hoping that the Jury would rule 'no' so that he wouldn't have to make a decision. That way he could of left it a gray area in the law that was up to interpretation. Now he has to do a lot of work and a lot of research and it's generally going to be a PITA. So from the Judge's perspective the Jury side of things was a bit of a bust. Now he has to do more work and that is irritating. The only thing worse then that would be to have his decision quickly overturned and then be made to look stupid.

Keep in mind that at this point they are mostly making it up as they go along. Also notice that copyright is unnatural, arbitrary, inherently conflicting, and designed with intentional unknowns and vast gray areas.
There is no basis in reality to go off here and is clearly very subjective.

Lets hope that the Judge will say that APIs are not copyrightable and end this wasteful farce right now.

Otherwise there are all sorts of other fun things to look forward to...

Unix copyrights + Linux + POSIX API
Microsoft copyright + Wine + Win32 API
Device firmwares + Drivers + Linux

And all sorts of fun stuff like that. The sky is the limit. Use a API created by another group? You would have some potential conflicts.

Google guilty of infringement in Oracle trial; future legal headaches loom(ars technica)

boog — Mon, 07 May 2012 21:53:24 +0000

The language around "assume" might seem ambiguous to a jury member (idiomatically, it could be interpreted as guidance that there was no controversy over the issue). However, if the judge goes on to rule that APIs can't be copyrighted, it will hardly matter that his instructions tended to help Oracle.

Google guilty of infringement in Oracle trial; future legal headaches loom(ars technica)

mikov — Mon, 07 May 2012 21:48:45 +0000

Fortunately being a lawyer is not a prerequisite for being able to form an informed opinion :-)

Google guilty of infringement in Oracle trial; future legal headaches loom(ars technica)

dskoll — Mon, 07 May 2012 21:20:14 +0000

Don't you agree that "assuming" that APIs are copyrigt-able, when Google's major defense is that they are not, further prejudices the jury against Google?

I don't know. I'm not a lawyer.

And what does "assume that X is true" mean, really?

I don't know. I'm not the judge. I can speculate that he wants to save time: He hasn't yet decided whether or not APIs are copyrightable, but if he decides they are, he doesn't want to have to go back and ask the jury whether or not Google violated the API copyright. But this is just speculation; you'd have to ask the judge.

Google guilty of infringement in Oracle trial; future legal headaches loom(ars technica)

mikov — Mon, 07 May 2012 21:11:30 +0000

Don't you agree that "assuming" that APIs are copyrigt-able, when Google's major defense is that they are not, further prejudices the jury against Google?

And what does "assume that X is true" mean, really? Jurors are not legal theorem provers that one can use to run different "what if" scenarios based on different input data :-)