Software Copyright and Innovation After Oracle v Google | Hacker News Reader

Software Copyright and Innovation After Oracle v Google

epiclaw.net

70 pointssean_hogle25 comments

This court case is, with no exaggeration, the most important software industry case in decades, perhaps ever. As the article says, if you're familiar with Android, Google, Chrome, Firefox, Edge (nee Internet Explorer), OS X, Linux, and the Internet, it is no exaggeration to say that every single one of those technologies, products, and companies are in violation of copyright according to this decision.

Oracle is literally trying to destroy innovation in the software industry, and they may just pull it off. Their products, their name, and their employees should be shunned for trying to destroy the single most important economic invention of all time: innovation in the software industry. This is not OK, this is not just politics, and they must be stopped.

Read this article, spread the news, write your congresspeople, donate to the EFF, and cancel your Oracle contracts.

The funny thing is that FSF/LPF actually pushed for such a boycott of Apple and a few other companies over the GUI copyright lawsuits during the 1988-1994 period. Needless to say, it wasn't effective and Apple etc eventually lost or dropped the lawsuit anyway.

> Oracle appealed this decision to the U.S. Court of Appeals for the Federal Circuit ("CAFC" or the "Federal Circuit"). Because Oracle's case included patent claims, the CAFC had jurisdiction over the appeal, and so Oracle was able to avoid review by the U.S. Court of Appeals for the Ninth Circuit

I think this needs to be changed. I can't think of any good reason why in a case like this, where it has both patent and copyright claims, that the appeals need to go to the same court. Why not have the copyright issues appealed to the Ninth Circuit and the patent issues appealed to the CAFC?

At trail court it makes sense to join issues that would normally be heard in separate courts together when they involve the same underlying facts, because there will be a lot of overlap in witnesses and evidence, and a trial court is both a trier of fact and a trier of law. Holding separate trials in different courts would be wasteful, annoying, and could get complicated and troubling if different courts found different facts.

Those considerations mostly do not apply at the appellate level. The appeals court does not determine facts. It does not hear witnesses. It just deals with the record from the trial court and the arguments from the attorneys.

> I think this needs to be changed. I can't think of any good reason why in a case like this, where it has both patent and copyright claims, that the appeals need to go to the same court.

Economy of justice concerns are the biggest reason, though there are other reasons to keep the case together (e.g., because it is one case between the same parties and different issues will rely on intersecting sets of determinations, so splitting it by the domain of the ultimate issues will open up the possibility of conflicting rulings on the same intermediate issues, which is not only ugly in and of itself, but especially ugly if the various parts of the split case are both appealed from the Court of Appeals to the Supreme Court.)

> Why not have the copyright issues appealed to the Ninth Circuit and the patent issues appealed to the CAFC?

I know the theory but has practice actually born out the supposed benefits of the CAFC as the single Court of Appeals for patent cases? Why not just send the whole thing to the Circuit Courts whether or not there are patent issues?

> At trial court it makes sense to join issues that would normally be heard in separate courts together when they involve the same underlying facts, because there will be a lot of overlap in witnesses and evidence, and a trial court is both a trier of fact and a trier of law. Holding separate trials in different courts would be wasteful, annoying, and could get complicated and troubling if different courts found different facts.

You really get all the same issues with splitting questions of law, since the ultimate legal decision on a patent question may depend on legal decisions on evidentiary and other questions also raised on appeal, on which ultimate non-patent counts in the same case may also rely. A clean separation is not realistic.

> The appeals court does not determine facts.

Yes, it does, though the standard of review for fact questions is usually one of "clear error".

> Why not just send the whole thing to the Circuit Courts whether or not there are patent issues?

Apparently, because previously different Circuit Courts interpreted the Patent Act differently, leading to forum shopping as problematic as what happens in ED Texas today. Except back then there was no recourse -- at least today you can take your appeals to a more consistent court.

> I know the theory but has practice actually born out the supposed benefits of the CAFC as the single Court of Appeals for patent cases?

This study partially tackles that question from multiple angles and the conclusion is mostly favorable to the existence of a CAFC:

http://papers.ssrn.com/sol3/papers.cfm?abstract_id=2231312

"The software industry, and indeed every industry that relies on software, has thrived for decades without the encumbrances of proprietary claims over APIs. Because the Federal Circuit's decision destroys the balance between copyrightable expression and uncopyrightable ideas in software, it threatens competition and innovation. The Ninth Circuit should repudiate it at the earliest opportunity."

BAM!

When presenting this to someone, is it helpful to use a human language analogy? For example: In english, when you refer to a "car", you are talking about a machine that moves people on roads. Whereas when you refer to a "boat", you are talking about something that floats on water. What Oracle wants Google to do, is to make up a new name for "car" and "boat", and try to get all developers to adopt these new definitions when using Google's products.

That doesn't sound very clear to me. I think a better analogy would be: Google and Oracle both made cars. Google put the gear shift in the same location as Oracle.

I think an even better analogy would be:

Google and Oracle and a bunch of other companies all make cars. All of the cars have a gearshift with the same pattern, because long ago the company who built the first car encouraged other carmakers to copy their design. But then Oracle buys the company who built the first car and tries to sue Google (and only Google) for copying their shift pattern.

Did Sun encourage others to copy the design though (genuinely asking)? They allowed others to use their designs. I have been under the impression that while Java was open-sourced, there was a a specific clause in the license to prevent what Google did.

Did Sun encourage others to copy the design though (genuinely asking)?

My understanding, which may be incorrect, is that Sun intended for their JVM to be the reference version, and they envisioned a lot of companies creating their own competing (but compatible) JVMs.

I don't recall any hints or suggestions "copying" the Java API would be a problem. I believe Sun planned on making money by charging for Java implementations to be tested against the TCK in order to earn the Java label.

I had read somewhere long time back that mobile devices were explicitly excluded from this in the license.

It's also interesting to read the counterpoint arguments of Florian Mueller on his FOSS patents blog.

http://www.fosspatents.com/2014/01/api-copyrightability-to-b...

The crucial point being that code written for Dalvik is actually not compatible with Java runtime. So how can you argue that borrowing code without consent is good for interoperability... when the product you deliver just break it?

The same Florian Mueller that never puts a disclaimer saying that he was an Oracle consultant for the case in the Oracle vs Google posts?

Thats unfair he put a lot of disclaimers in his post. It's just that they don't appears in the TL/DR that people share afterwards :P

I explained it to my parents by saying "You can copyright your street address, and anyone who writes it on an envelope to be mailed to you has just committed a copyright violation."

They said that sounded pretty stupid, I agreed, but its the current law as far as I can tell.

Could you explain more please.

Making up a street address would normally not involve creative expression, but even if it does, its intended purpose is to allow others to use the address for various purposes, which would then be fair use.

In this analogy, Google did not mail something to this address, rather created another house with the same address so that anyone sending mail to Oracle is able to redirect it to Google without changing the address (by posting into a different mailbox).

Design of an API certainly involves creative expression, so much that many developers are not able to get it right. So I do not think the analogy is valid anyways.

The basic analogy is that an api is like the pointy end of a drill bit. Google and Oracle want to help people drill in screws or hex keys. Sun started selling flat screw bits, and google, in order for handymen (developers) to be able to use the same skills, also started selling flat bits (or flat head screws if you prefer).

The one distinction complicating this a little is more thought and creativity goes into designing an api then designing the interface between a drill bit and screw.

This is a risk of using proprietary languages. Even if the implementation is open source if a single company actually controls the language you are running risks. In addition with Java there is the precedent of Sun and Visual J++ so it was known that Sun would go after people who provided incompatible implementations. This whole saga is simply a very bad move on Google's part with foreseeable consequences.

The difference is that Microsoft called the incompatible implementation Java.

Google never called it Java

I don't see a significant difference. If you look at the Google Android developer page http://developer.android.com/guide/components/fundamentals.h...

The first sentence is "Android apps are written in the Java programming language."

If that is not calling it Java, it is close enough as to be indistinguishable to the average programmer.

Yes, the programming language, not the platform. Android doesn't run in a Java Virtual Machine, it only uses the Java syntax

Microsoft was sued because they certified the platform and it was incompatible.

Google/Apache never certified the platform to call it Java

The Sega-vs-Accolade times seem so far away ... https://en.wikipedia.org/wiki/Sega_v._Accolade

The thing I will never understand about the case is why the one accused is Google when they used the Apache Harmony implementation done by the Apache Foundation.

Well reasoned. I still cannot understand how the court chose to ignore the unusually clear precedent of Lotus v. Borland. That took some contortions.

Reply on news.ycombinator.com