Read on Twitter
Google v. Oracle, Supreme Court, Oct. 7, 2020. CSPAN livestream here, it's also on CSPAN radio:
https://www.c-span.org/video/?469263-1/google-v-oracle-america-oral-argument
https://www.c-span.org/video/?469263-1/google-v-oracle-america-oral-argument
love me my teleconference oyez oyez oyez
I wrote a refresher yesterday: https://www.theverge.com/2020/10/6/21504715/google-v-oracle-supreme-court-hearings-android-java
Google is up first. Roberts has the first question, asking it would be "okay" if someone took a brief and all the headings and their organization, so long as the copier rewrote the brief.
Google says no, but says it's different, which is WEIRD, imo, because doesn't it depend on the headers and the nature of the headers??? But um also then there's the part with the copyrightability legal briefs lol
Roberts: "cracking the safe might be the only way to get the money you want, but that doesn't mean you can do it. If it's the only way, the way for you to do it is to get a license"
we are not onto a good start
we are not onto a good start
Google says that if you patent the safe, you can keep people out, but if you write a book about how to crack the safe, that doesn't mean you can keep people out of it. Pretty good response tbh. We're onto Thomas now
Oh god
Oh god what
Thomas asks why they aren't relying on Sec 101 more than Sec 102(b), which I assume is a question about CONTU, but _why_
Now discussing Baker v. Selden, an 1879 case about a system of bookkeeping. If you're a computer programmer interested in this case and haven't heard of Baker v. Selden, you're in for a treat https://en.wikipedia.org/wiki/Baker_v._Selden
I can't have heard this right, but I thought I heard oralist for Google say, "Oracle doesn't want a fanbase, Oracle wants prisoners."
Onto Breyer, the justice who tends to be most friendly to fair use. "Say I put java.lang.Math.max into my computer..."
Asks why you can't write a computer program that does the same thing without using the exact formulation java.lang.Math.max
Asks why you can't write a computer program that does the same thing without using the exact formulation java.lang.Math.max
Google has a nuanced reply but "Why would we have a copyright system that does that" is honestly my favorite nutshell response
Alito up now asking, "You're arguing in a circle saying there is only one way to write the code."
Google: "We would happily not use the Java SE declarations if we could, but the language only permits us to use those."
Sotomayor: You agree could not have copied your implementing code, but you fight the declaring codes, there's ways to write your own declaring code as well. Apple did that. (aka Apple's Swift)
Google: With respect to Apple, it's true it didn't reuse the Java SE declarations because it wasn't using Java. But it did use other declarations.
(Swift is descended from Objective-C, Objective-C is descended from C)
(Swift is descended from Objective-C, Objective-C is descended from C)
Sotomayor: What gives you the right to use their original work?
...
My problem with your argument is, what's your definition of interoperability?
...
My problem with your argument is, what's your definition of interoperability?
Sotomayor: You've now developed many different software packages now... can anyone copy those now?
Google: ... they are certainly free to copy the interfaces. ... Incredibly common for someone to ... take the functionality of the computer program and do it better.
Google: ... they are certainly free to copy the interfaces. ... Incredibly common for someone to ... take the functionality of the computer program and do it better.
Kagan: I have to confess to being a little surprised or confused by some of the arguments you're making this morning.
Kagan: I took you to be saying [in your brief] that the the declaring code is unprotectable because it's a method of operation. I don't hear you saying that today. I hear you saying that the real question is, are there multiple ways of doing the same thing?
Google: They're different arguments. We're making both of them.
Kagan: If that's your test, is that the test that comes out of the 2nd circuit Altai case? Is there any difference between what you say and what Altai says — which is that we have to separate out the expressive elements of something?
Kagan: Suppose I'm sitting in a mathematics class and someone says do a proof, and it turns out 20 people in this mathematics class come up with more than one proof, and some are better than others. How do we deal with that? I would argue that's pretty analogous to the situation
Kagan: There's more than one way, and Oracle came up with the better one.
(Historical note: Oracle acquired Sun in 2010, and with it, Java).
(It is generally thought that Sun would have never sued Google, even if there was bad blood over the reimplementation of Java without a license).
(Incidentally, mathematical proofs are not copyrightable material)
Onto Gorsuch. "I'm stuck in a similar place to Justice Kagan [about the merger doctrine]. ... Why ride on the innovation of a competitor that comes up with a more successful solution?"
Gorsuch: Why has your competitor [Apple] been able to come up with phones that work fine without copying?
Gorsuch: If we're worried about ideas and expressions merging and others have been able to accomplish the task without reliance on what you claim to be the essential facility, where do we stand?
Google: If the test is a high level generality, the idea of being able to create a phone, fine, but that's not the test. The test is to look at the copyrighted work and the expression.
Onto Kavanaugh.
Google: Declaring code... tells the outside developer what to do. The developer looks at the declaring code and then knows how to operate the short-term prewritten programs. It tells someone else how to operate the computer programs. That is unlikely any other code.
Kavanaugh: You're not allowed to copy a song because it's the only way to express that song. Why is that principle not at play here?
Google differentiates between a song and a work (like the accounting book in Baker v. Selden) that tells the public how to do something.
Google differentiates between a song and a work (like the accounting book in Baker v. Selden) that tells the public how to do something.
Kavanaugh: Not aware that the sky has fallen in the last 5-6 years with the Federal Circuit 2014 ruling on the books.
Google: We won the fair use trial [after]. ... I would not then say that the representations of the country's leading computer scientists and the computer industry are false
Google wraps up: The declarations so lack creativity that they deserve the least copyright protection. By claiming the exclusive right to the declaration's function, Oracle is asserting a patent right to insulate itself from competition.
Oracle is beginning.
Oracle says a handful of legal principles govern this case. Legal principle one, Congress granted copyright protection to software.
Oracle calls Android "a commercial sequel that meets the purpose and market of the original."
"If this court holds that a jury may conclude that copying declaring code is fair, it will create uncertainty and decimate a business model that companies rely upon."
Roberts compares to the organization of a restaurant menu — appetizer, entree, dessert. "Every restaurant menu organizes their menu that way."
Oracle: There's scene-a-faire. If there's a standard way of doing things, it's not protected. What we've got here is very different. It's not just, here are apps and here are dinner plates. We fill the blanks in.
(Oracle is saying scene-a-faire is different from merger doctrine, and also doesn't apply).
Oracle: Microsoft and Apple both spent billions of dollars creating their competing platforms. Copyright Act does not give a free pass just because it would be expensive to create your own expression.
Oracle: No other context would it be appropriate to be asking whether the work is unprotected or whether there is fair use by saying that the audience has learned the words by heart.
Thomas: You're arguing for more than the declaring code. If they are right, do we need to decide more than that?
Oracle: No, all this court has to decide is whether the declaring code for the purposes was original. It was.
Oracle: No, all this court has to decide is whether the declaring code for the purposes was original. It was.
Oracle: You cannot distinguish declaring code from implementing code in the way that Congress intended.
oh my god
oh my god
oh my god oh no
Oracle: Congress rejected the exact line that Google proposed in Sec 101.
Thomas: What would in your way "transformative" look like for computer code?
Thomas: What would in your way "transformative" look like for computer code?
Oracle: Adapting our code into a smartphone is no different from adapting a book into a movie, a superseding commercial market.
Breyer: Assume with me the following. Assume that computer programs, which do something after all, are copyrightable. But methods of operations are not. The problem for us is, is this like Baker v. Selden, or more like an ordinary computer program?
Breyer compares to the QWERTY keyboard. If someone had a copyright on QWERTY, they'd control all the keyboards.
Compare also to phone switchboard and to a spice recipe. )All of these are the subject of patent and not copyright)
Oracle: This is not like the QWERTY keyboard, there was never anything expressive in QWERTY.
Alito asks about standard of review and civil procedure, which frankly I thought would wind up dominating today's argument, but here we are in metaphor hell instead
(Historical note: Oracle did bring patent claims in the original trial, and Google won on all of them).
Alito asks if it was an error to bring the fair use case to a jury. Oracle says no, but says that the standard of review was correct, to have the Federal Circuit review de novo
Sotomayor: You had the sky falling if we ruled in favor of Google. It seems that since 1992, a basic principle has developed in the case law, up until the Fed Cir decision, I know there was a 3rd Circuit case in the 1980s, but they and the other courts have moved away from it
Sotomayor: Industries have grown up around the principle that they can [reimplement].
Sotomayor says that whether it's Apple's desktop or AWS, industry thinks declaring code is not copyrightable, but implementing code is. "Why should we change that understanding?"
Oracle: Not a single case has ever said you can copy that amount of code onto a distinct platform for that purpose.
Oracle: No one drew that distinction between declaring code and implementing code.
(The cases that Sotomayor reference generally fall under the umbrella of reverse engineering software)
Kagan asks if the declaring code is divided into "the trigger" and "the method of organization," and one is copyrightable and the other not.
Oracle now, for the first time, brings up "structure, sequence, and organization"
Oracle now, for the first time, brings up "structure, sequence, and organization"
Kagan: Suppose I own a grocery store, and I come up with a great way of organizing all my fresh produce into categories, very intuitive for the shopper
"Do I have a copyright claim?"
Oracle: You have a copyright claim in anything you have set down in writing... maybe? But this is worlds different —
Oracle: You have a copyright claim in anything you have set down in writing... maybe? But this is worlds different —
Kagan asks why not. Compares to the kingdom/phylla organization in biology. "There are a thousand ways of organizing things, which the first person who developed them, they could have a copyright."
Oracle: It's not just the most intricate hierarchy you've ever seen... the relationships cross from one package to the next, from one class to the next. It's extraordinarily intricate in a way that does deserve copyright protection.
Gorsuch back on the Alito train with the standard of review questions
(Historical note: The QWERTY keyboard argument featured *heavily* in Google's briefs in the first Federal Circuit appeal, and Google lost then. It's interesting to see it come back and taken seriously)
Kavanaugh following up on QWERTY and Sotomayor's question about settled expectations and the 83 Computer Scientists' amicus brief.
Pretty open-ended question.
Oracle: This case would be like Baker if we were trying to block others from using their own package class method structure to overwrite their own prewritten programs
Oracle: We have heard dire predictions from Google about software innovation. The software industry rose to world domination because copyright protection, not unlicensed copying.
Oracle: In that time we've seen the explosion of interoperability.*
*NARRATOR: We did not.
Oracle takes a minute to wrap up. "Ruling for Google will decimate the incentive to create high-quality user-facing declaring code. Code that the amicii on both sides is essential for the industry to survive. That will hurt app developers and industry in the long run."
Says big companies are paying to license declaring code
"Our rule protects original code, it's a simple rule, it comports with copyright principles."
Deputy Solicitor General now has the podium. (Yes, the government is getting a turn in this case).
"Google's core argument is that it would be inefficient to make them learn new calls in order to invoke new declarations..." But says it's always more efficient to copy a work anyways.
Roberts: Google says if we agree with Oracle, we will ruin the tech industry in the United States. What do you have to say about that?
First mention of open-source licensing in this entire argument.
SG says tech industry will not be adversely impacted by agreeing with Oracle.
lol
SG describes the standard of review as: "We'll assume the version of the facts most in Google's favor, but we will determine as a matter of law, was this use transformative"
SG: The court should also be asking, what incentives to future innovation should the rule create
Breyer: Why do you think the balance of harms lies the way it does?
back to the qwerty
back to the qwerty
We are now a little over an hour into oral argument and I'm incredibly depressed with how this is going
SG says declaring code and "incredibly intricate" SSO are sufficiently creative to merit copyright protection
Alito: There's an argument that the sky is going to fall if we don't rule for Google. Do you have anything you want to add on that point?
Stewart: There is a phenomenon of licensed copying.
Explaining open source now
Explaining open source now
Reading the tea leaves, it seems like the court is weighing SG's opinion on how this affects the tech industry more heavily than what the tech industry has to say about itself
Sotomayor asks why this isn't fair use.
SG says it's not transformative to livestream a cam of a movie theater release
Kagan: When a smartphone manufacturer uses the QWERTY keyboard layout and uses it for its next phone, is that fair use or not?
[assuming QWERTY is copyrightable]
SG: We don't have a quarrel with a proposition that preserving interoperability is favorable to a fair use analysis.
Kagan: So why is that not the case here?
Kagan: So why is that not the case here?
SG says app developers are in a fundamentally different position than consumers.
I am not sure I follow the logic here
SG says that app developers are "independent economic actors whose interests align with Google's"
Gorsuch asks why no reasonable jury could have concluded it was fair use here. "Aren't you essentially saying that yes, code is copyrightable, but it's always subject to fair use?"
SG says that it should have been a subsidiary legal judgment and not a factual judgment. "is this sufficient for a transformative use" was a legal issue and therefore for a judge
Says this standard of review issue will affect summary judgment practice (not sure why it would)
Kavanaugh: Google says in its reply brief that the undisputed fact in this case is that the declarations could not be written in any other way and still properly respond to Java programmers. Are they wrong in saying that?
SG does not dispute, but now we're into the philosophical weeds of at what ontological point the merger doctrine begins to apply
SG now discussing CONTU
SG wrapping up.
Google is getting another round of questions.
Roberts asks if there's further response to Oracle's representation about effects on tech market.
Google: I don't think Mr. Stewart [SG] is accurately respresenting the effect.
Google: I don't think Mr. Stewart [SG] is accurately respresenting the effect.
Google says that deferring to the jury is important because it's so fact-bound.
Roberts asking why merger argument doesn't make Oracle a "victim of its own success."
Asks why Google couldn't just do the same thing as its competitors.
Google: I think that's an optical illusion. Both Apple and Microsoft did reimplement prior interfaces. The reason why they didn't use *these* interfaces is because they were writing in a different language.
Google: I think that's an optical illusion. Both Apple and Microsoft did reimplement prior interfaces. The reason why they didn't use *these* interfaces is because they were writing in a different language.
Honestly this is the most persuasive thing in this case and it's gotten four seconds in the oral argument
Breyer: Copyright runs from the beginning. What do you do about that?
Here we are in the ontology of the merger doctrine again
Here we are in the ontology of the merger doctrine again
Google: Our best argument is about the standard of review. This case was put to the jury at Oracle's insistence.
Kagan asking if transformative use is the right question here. "As I understand it, you're using this for the exact same purpose.... is the transformative use really a mismatch in this purpose?"
Google says interface is "barely creative," it is connective tissue. (That changes how you weigh transformativeness)
Gorsuch asks about negatively impacting summary judgment practice.
Google: Courts have had no problem reaching summary judgment where it's appropriate... classical fair use cases are things like parodies or news reporting where we have established rules.
Kavanaugh asks if Google has more to say about the merger doctrine.
Google asks rhetorically why Congress would want a rule that makes it as inefficient as possible for developers.
Google now wraps up, focusing on jury vs. judge and standard of review
Google now wraps up, focusing on jury vs. judge and standard of review
Google: Jury heard testimony on a variety of points that Mr. Rosencranz [Oracle] is trying to deny.
...
There is a widespread consensus in the industry and among computer scientists that this was the practice. Why would we want to upend that?
...
There is a widespread consensus in the industry and among computer scientists that this was the practice. Why would we want to upend that?
Says declaring code is "minimally expressive."
Federal Circuit deemed jury verdict "advisory."
Google: That is not appropriate. It is impossible to unpack the factual findings they were relying on.
Google: That is not appropriate. It is impossible to unpack the factual findings they were relying on.
Google goes on a pretty excellent spree of picking apart Oracle's representations just now with rebuttals from the district court trial (including Sun CEO testimony), saying that these are very well what the jury was relying on
Oral argument has wrapped up.
My impressions:
Roberts: I have a generally poor opinion of how Roberts deals with complex technical issues and this has not helped, honestly surprised he didn't just stick to the standard of review
Thomas: primarily interested in CONTU, for some reason
Roberts: I have a generally poor opinion of how Roberts deals with complex technical issues and this has not helped, honestly surprised he didn't just stick to the standard of review
Thomas: primarily interested in CONTU, for some reason
Breyer: persuaded by QWERTY
Alito: One foot in "this is not how merger works" land, the other foot in standard of review land
Sotomayor: persuaded by QWERTY
Kagan: somewhat persuaded by QWERTY but prone to thinking along the same lines as Alito
Alito: One foot in "this is not how merger works" land, the other foot in standard of review land
Sotomayor: persuaded by QWERTY
Kagan: somewhat persuaded by QWERTY but prone to thinking along the same lines as Alito
Gorsuch: Wants to decide on standard of review
Kavanaugh: as junior justice, always goes last in a teleconference argument (it's hard to signal the classical interrupt-free-for-all without the visual cues) and his questions were a little thin on clues as to what he's thinking
Kavanaugh: as junior justice, always goes last in a teleconference argument (it's hard to signal the classical interrupt-free-for-all without the visual cues) and his questions were a little thin on clues as to what he's thinking
And not a single thread of textualism in the bunch! Given the text of 102(b) I would have thought it would have come up.
I am not clear on how Gorsuch, Alito, and Kavanaugh feel about standard of review, someone who is more tuned into Supreme Court dynamics and civil procedure should feel free to chime in here.
I truly did not expect Thomas to be pushing to decide this on the basis of CONTU for some reason but in a way, that is the most Thomas take on Oracle v. Google that could be possible
I don't like to make predictions about how courts are going to rule, especially the Supreme Court, but I am not optimistic about Google's chances here.
I am extremely disturbed at how the court discussed the nature of declaring code and additionally disturbed by representations and insinuations made about Swift, Objective-C, and so forth.
These are facts that were not in evidence at the trial, and furthermore, a closer look at all of these languages is not actually favorable to Oracle's position.
I do not think Oracle would have allowed (nor should they have allowed) detailed evidence at trial of wider industry practices. For example, comparisons between, say, SQL and mySQL.
In that respect, to now be saying that Google could have just done what Apple had done (aka, base Android development on language that's *not* Java or even derived from Java) is certainly an optical illusion.
tbf I don't know what any of this is either. love you guys https://twitter.com/struberg/status/1313869622559543297
None of the comparisons are particularly good, it's been a consistent problem with the case. https://twitter.com/codepaintsleep/status/1313873180906455040
I wrote about it a little here https://www.vice.com/en/article/mg77dy/in-google-v-oracle-the-nerds-are-getting-owned
(to be clear that's from the second trial in 2016. up top earlier in the thread I wrote a present-day refresher).
Just thought about the irony of the SG being apparently concerned about how this case would affect summary judgment practice. Oh, you think a bad decision might make it....... harder to do your job? 
This is the first I've considered the idea that the saga of Oracle v. Google will end with legislation, but it very well may. https://twitter.com/charles_duan/status/1313879330355924992
I keep spinning out some of the consequences in my head and mostly it's chaos
But if it's any consolation, even if you're a software developer — heck, even if you're a software developer who reimplements APIs — Oracle v. Google won't be the worst thing that happened to you in 2020
suck it up, buttercups, what's another trashcan fire on top of the big ol flamin dumpster we've already got going
Maybe this wouldn’t have been tactful but it’s interesting that Google didn’t respond to Kavanaugh’s question about the sky falling, pointing out that the Federal Circuit doesn’t have jurisdiction over software copyright lawsuits,
so the “settled law” in favor of Oracle since 2018 (arguably since 2014) only exists in one jurisdiction that you can only get to by asserting a patent claim at the start
how are you supposed to measure the effects on an industry for the federal circuit’s interpretation of 9th circuit law, pending SCOTUS review
you sure as shit will get to measure it once there’s SCOTUS precedent, though.
I know it’s been hours now but I keep remembering new things to be horrified by, like how no one seemed to focus on the delineation between patent and copyright, and how all their examples kept veering into patent land
Some of it was rhetorical, yes! But the safe-cracking example that arguments opened with was astonishingly bad and deeply confusing to boot
A decision is published sometime next year. A lot of outcomes are possible — but if Oracle wins and the Fed Cir 2018 ruling stands, case goes back to San Francisco for another trial to calculate damages. What the industry does in response, ¯\\_(ツ)_/¯ https://twitter.com/bluecmd/status/1313912687089377280
Yes — and beyond that, what about how the IRAC format is standard? https://twitter.com/jefatlaw/status/1313912125035749376
It’s true that the entire bench is filled with humanities majors, but the legal profession — *especially* the rareified ecosystem of the Supreme Court — deals regularly in highly ritualized human-to-human communications that are required to arbitrarily follow specific formats
here’s the thing about this case — I am not a programmer and never was. But my extremely rudimentary education in programming is enough to alarm me when lawyers claim that declaring code is a creative, copyrightable expression.
you know a case is bad news when you get into a casual conversation about it that devolves into “but what is a computer language, *really*?”
but for real, instead of dropping the 2736th analogy about hamburgers into my mentions, how about philosophizing on the essential nature of a computer language for a change
The official transcript in the Supreme Court oral argument for Google v. Oracle is now online, so you don't have to take my word for what was said! https://www.supremecourt.gov/oral_arguments/argument_transcripts/2020/18-956_kifl.pdf
Breyer was being rhetorical here but this still makes me want to shoot myself into the sun
launch me into the sun and then launch the sun into a larger sun
the entirety of java.lang, bud https://twitter.com/rjkucia/status/1313946011824521226
(this is the part where people start tearing their hair out)
maybe I will be pleasantly surprised and the final decision will have no resemblance to any single portion of this oral argument (it happens! sometimes!) but you'd think there's a single portion of the oral argument where i'd go "yes, I think this is the part that makes sense"
