Oral argument — Apple Inc. v. Samsung Electronics Co., Ltd.

The first case for argument this morning is 15-1171, Apple v. Samsung. Ms. Sullivan, I'm going to introduce you. Good morning, Your Honors, and may it please the Court. Kathleen Sullivan for Samsung. I'd like to begin with why the Court should reverse on the 647 Quick Links patent. And to begin, I'd like to return to the key language in the claim as construed by this Court in Motorola. The claim requires an analyzer server for detecting structures in the data and for linking actions to the detected structure. So what we need is an analyzer server for linking and detecting. But as this Court held in the Motorola case, construing the exact same patent claim, analyzer server must be separate from the client application that is sending the data to the server. And the exact words at Slip Opinion 23 in Apple v. Motorola is that the specification requires an analyzer server that is separate from the application of the server. The term server was not construed in this case, was it? That's correct, Your Honor. But in the Motorola case, we did look at the term server. You did? Yes, and said that the plain meaning of that term was that it required a client-server relationship. Is that correct? That's exactly right, Your Honor. And the exact words, if I could just recall them, are you construed analyzer server as, quote, a server routine separate from a client that receives data having structures from the client. And, in fact, you pointed at Slip Op 23 to Figure 1 in the patent, which has a box for the application, Box 167, and a box for the server, Box 165. And you said the specification describes the analyzer server and the application which it serves as separate structures. So I want to focus on separateness and show why, very simply, there is no separateness on the record here. And, therefore, analyzer server can't be met. But I guess the question here, as you say, is what does separate mean? And if I understand the situation, Apple's expert admitted that the library program was not a stand-alone program. Yes, Your Honor. But it is also not something that's routinely incorporated into the client program. It's called up when necessary, correct? Your Honor, you're getting to what is applicable for detecting. I wonder if I could start with linking, which doesn't involve the shared library at all. Could I start with linking and then address what happens with the shared library? Because shared library, there's an answer to your question, Your Honor. But what may not have been entirely clear in the briefing, but I want to make utterly clear to you, is that if we focus just on linking, we don't need to get to shared libraries. And what I'd like to do is go back to Apple's expert's admissions about the linking functions. The linking functions need to be in the server, not in the application. But we have two moments in time before and after this court's Motorola construction. We have the moment in time where first Apple's expert, Dr. Mowrey, comes in and he says, oh, the linking code. I'll tell you what the linking code is in Messenger and browser. And this is at A10857, 10859, and 10863 in the record. And he says there are three routines that do the linking. And they are add call and contact menu, show links context menu, and create context menu. Now, these are not in the shared library, Your Honor. These are in the application. And at that time, Dr. Mowrey says, those codes, that code is in the application. And why, Your Honor? Because you haven't told him yet that you reject his construction. You haven't told him yet that he can't just say the server can be any routine. And remember, Apple's pre-Motorola construction of 647 was, they said, an analyzer server should be construed as a program routine, a program routine that receives data. And you said, no, it can't be any program routine. It has to be a program routine that is separate from the client. So, so far, Your Honor, pre-Motorola, Mowrey is saying that the linking code is in the client. Now, what happens after Motorola comes down? Motorola comes down at the 11th hour in the trial, and Judge Koh allows the parties to come back, bring the experts back, Dr. Mowrey comes back. And if you look at nothing else in the record, Your Honors, please look at pages A13033 to A13034, because that's what Dr. Mowrey says about the linking code when he comes back. And here's what he says about the linking code. He says that 13033 lines 8 and 20, well, the code that does the linking, it's still those three routines I told you about earlier. It's add calling contact menu, it's show links context menu, and it's on create context menu. And he still says those three routines are in the client. Your Honor, look at A13033 line 10 and A13034 line 18. He says those three routines that do the linking are in the client or from the client. They're in browser or they're in messenger. And now comes the key moment, Your Honor, and here's where Apple is trying to end run this court's construction of analyzer server as being separate from the client. He says, oh, well, they're in the client, and they're from the client, those three routines that do the linking. But never mind, they're just glue code. They're just glue code. Now, I don't care what he renames them. He can rename them glue code. He can rename them analyzer server. He can rename them server routine. But he can't meet the construction this court gave to analyzer server unless they're outside the client. Linking code that's in the client cannot be functioning as an analyzer server which has to be outside the client. So this is the key, Your Honor. This is dispositive. You don't have to get to shared libraries, because we're done right there. But I hope you're going to get there. I will, Your Honor. I very much want to answer Your Honor's question. I just wanted to be clear, because it may not have been clear in the briefs, that as to linking, you never need to get to shared libraries. The linking code is entirely a matter outside the shared libraries, because Apple admits that to link, you need code that's in the application. Now, Your Honor, thank you for your patience. And I do want to get to shared libraries and say why we also went on shared libraries. Now, Dr. Malley admitted that— The question is what does separate mean? There's no question that it's a separate program in the sense that it's not routinely run as part of the client program. But as I understand, Apple's expert also admitted that it's not run as a standalone program. That's correct, Your Honor. And does separateness require that it be run as a standalone program, or is it sufficient that it's a separate program that's called up and run by the client program when necessary? The former and not the latter, Your Honor. It must run as a standalone program separate from the client. If the client calls up the program and incorporates it for purposes of performing linking and detecting, it is not an analyzer server, and here's why. The claim, just to go back to the language of the claim, the claim requires a server routine separate from a client that receives data having structures from the client. That's your construction. I'm sorry, I want to go back to the claim language. Claim one requires an analyzer server for detecting and linking. And I think the plain language there means that the code that is— the analyzer server has to be doing the detecting and the linking. And the analyzer server is not doing the detecting and the linking. The application is doing the detecting and linking, incorporating or importing the code from the shared library to do that. And the key—admission is really one that my friend, Mr. Lee, emphasizes in his brief at page 17. If you want to see the key admission as to why it's the application that's doing the detecting and linking, it's Dr. Mary's own words at the bottom of red brief 17. This is a quote from the record at page 13036 to 13037. And this is a key sentence, and it's emphasized by Apple, but I think it—with all respect, it cuts our way. Well, it has access to the code, and it goes to the code where it is and uses it there, and it does that each time that it accesses the code. Well, the it in the beginning of that sentence that is the subject of the verb uses is the application. The application is using the shared library code, meaning it's the application that is doing the detecting and linking. And it imports. If you look at the source code, the jury has the source code. It's JX50. There are commands that say import. So there I am in Messenger. There I am in Browser, and I say import the link of five files I need or import the other files I need to do the detecting. But, Your Honor, if I could use a humble real-world analogy, what your construction wants is the following. If I have to perform a calculation, you want me to send the equation to the reference librarian and have her calculate it for me and send it back. That's why you said at slipout 23 that the application in box 167 is calling on the program, the separate structure in 165. So if I send my equation to the reference librarian and I say, please calculate it and send it back to me, that's your construction. But what Dr. Mowery has testified here to Your Honor is I go to the library, to use my humble analogy, and I take a book out and I import its knowledge into my head and I perform the calculation myself. So I think once you focus on the key language in Claim 1, that it is an analyzer server for detecting and linking, and once you look at Mowery's admission that it is the application that is using the shared library code to perform detecting and linking, and you add that, Your Honor, to the key point you made earlier. You focused on Dr. Mowery's admission that the library code cannot function alone. And again, if you want to look at the key page, it's 13054, where Dr. Mowery says, using Linkify as an example, as it is running, it is not running as a standalone program. So Your Honor, to answer your question more succinctly, separate means that the analyzer server has to be separately, independently, sufficiently as a standalone program doing the detecting and linking, just as if the reference librarian did my calculation for me. Your Honor, I'm concerned about time and I want to make sure I get to some other patents, but I want to make sure I answer your questions on 647. I have a question with respect to the 5-9 patent. And this is the J-Mall of Invalidity. What evidence was introduced with respect to defining or limiting the term heuristic? So Your Honor, the evidence introduced about the meaning of that came in, in part, from inventor testimony. If you want to see why we think heuristics is indefinite, the key place to go would be to the page 47 of the yellow brief. And we cite there to... It seems to me that the use of the term, with nothing else, would indicate perhaps indefiniteness. So I'm asking you to point out, and I'll ask the same of Mr. Lee, to point out what evidence was introduced that defines or limits what heuristic means here. So the evidence was partly inventor testimony, but the inventor testimony shows why it's indefinite, because the inventors themselves contradicted each other. We don't dispute that the testimony, the evidence, enabled heuristic to be construed. What we contest is whether the evidence allows heuristics to be construed with the definiteness or boundaries required by Nautilus. And while the district court, of course, took Nautilus into account, we don't think that this use of heuristics is definite. But Your Honor, you don't need to reach that, because we think you can affirm the judgment of non-infringement of the 959 patent. Wait a minute. What do you mean we don't need to reach that? I thought that you were arguing that we did need to reach it, and that you thought the patent claim should be invalid, even if we were to sustain the non-infringement judgment. Your Honor, we did raise invalidity as a counterclaim, and so we did give you the opportunity to invalidate the patent. Opportunity? Are you— Suppose, hypothetically, we were to agree that there was no infringement of these claims. Do we need to reach the invalidity issue? You do not, Your Honor, because you could affirm the judgment here without reaching it. Okay. And we think the case for non-infringement, the substantial evidence supporting the judgment of non-infringement, is overwhelming. And I won't go into that, because that's in the brief. Your Honor, if— Can we move to 721? Yes, Your Honor. 721. Yes, Your Honor. Okay. Let's assume, for the sake of argument, that we have some disagreement with the teach-away conclusion that the district court reached and, arguably, the jury reached. Yes, Your Honor. If we're left with only the secondary considerations, why are they not sufficient? One, what standard of review do we apply? And two, why are they not sufficient basis for us to sustain the finding of non-obviousness? Your Honor, we think it's a legal question. We think that the obviousness is a legal question, and the ultimate conclusion is one that you can review de novo. And, therefore, the question of whether the secondary considerations are strong enough to overwhelm a strong case of obviousness is a question that you can arrive at without deference to the district court. But, with respect, we think, second, it's a legal question whether Plazant teaches away. And I think that it's obvious— Let's leave that as a teach-away. We don't have time for that. Justice, let's end there. We— Are the strong secondary considerations, are they not particularly given the comments by Samsung's own employees with regard to the ingenuity of this? Or I forget the asset tips they used. So, they're not for several reasons, Your Honor. First, there was no notice of the 721 patent before the suit. So, any observations about general sliding to unlock is not relevant to the commercial success of the particular slide to unlock claim here. So, the other thing— Yeah, but they were talking about the features in the Apple device, right? Yes, Your Honor. There is evidence that that— The body of the patent, even though the patent might not have been noticed at that time. Yes, Your Honor. But any evidence of copying predates the patent. It's pre-2010. And so, again, I just want to focus on what we're asking here is not whether features or concepts— Why isn't copying relevant even if it occurred before the issuance of the patent? Well, Your Honor, what you're looking for in secondary indicia is, is this the reason why there was commercial success? And the evidence here was very weak as to commercial success. Copying separate from commercial success, right? Yes, but the copying— So, I guess what I'm trying to suggest, Your Honor, is that the level of particularity at which you should care about copying is, was the patent copied? Was this claim copied? Was this particular iteration of slide-to-unlock copied? I don't understand that. It seems to me the question is whether the feature was copied, and if it was copied even before the patent issued, that there's still into a secondary—a relevant secondary consideration. Your Honor, I'd submit at most it's extraordinarily weak because it's too general. But let me give you two other key points about commercial success, if I could. Apple doesn't practice this patent in its new operating system, in iOS 7. But let's forget about commercial success. What about long-felt need? What about industry praise? I mean, commercial success isn't the only thing we look at. Fair enough, Your Honor, but there's no evidence of nexus between this patent claim and commercial success or the other factors. The Hauser study didn't test the 721 on smartphones. It tested it on tablets. The long-felt need, Your Honor, is referenced in both the prior art references. It's referenced in Neonode and Plaisance. The long-felt need to avoid unanticipated activation is in the prior art. It's not something that Apple introduces to the world. It's a long-felt need that's met by other— What about the Condroit study? That went in evidence on damages, right? That wasn't evidence presented in connection with this obviousness question. Well, Your Honor, Apple claims that it's evidence of commercial success. No, I know that they used to do. We think that you should reject it. I mean, we think that the study is flawed for obvious reasons. We've discussed it before. But the key point, Your Honor, is it didn't test 721 on phones, and therefore it isn't evidence of commercial success. And there's no showing of commercial success of this feature in particular on phones as opposed to the iPhone more generally. So, Your Honor, with respect to copying evidence, even if I concede that copying a feature—we don't concede that we copied features, but even if that were relevant to this inquiry, it's too soon and it's not relevant to this claimed feature in the patent. There are lots of ways to slide and unlock. Neonode and Plaisant anticipated them, and we think a person with skill in the arts would have known to combine them. We put in expert testimony that was powerful on that, and Apple didn't put anything to the contrary. So, Your Honor, to answer your question, it's a way. You have to look at the secondary indicia, which we think are very weak, and weigh it against the case for obviousness, which we think is very strong. And we think that coming out on that as a matter of law, you should find the 721 invalid as obvious. And one last factor, it is worth noting, although obviously we don't have to defer to other nations of the world, all the other jurisdictions of the world who have considered the 721 patent or its foreign equivalent in light of these prior art references have invalidated it. So, 721 has been invalidated elsewhere in the world based on obviousness from these references. Your Honor, I'm... 172. 172... Why do you want to sit down? Well, Your Honor, I want to reserve time for... Why do you want everyone to sit down? Well, Your Honor, will I have time to rebuttal afterwards? Yes. I certainly want to answer all of your questions. So, Your Honor, as to 172... The secondary considerations, in fact, were the 172. Same argument there, Your Honor. And, again, I'm asking you to please consider secondary considerations in relationship to the strength of the obviousness case as a matter of law. There may be implicit factual findings, but you have the ultimate decision about how to weigh them. On the 172, Your Honor, we think there's, again, a very strong case for obviousness. We think Robinson is almost a single-reference obviousness reference because it disposes every element of autocorrect except for one thing, the text being typed in the first area. And we think it was well-known in the arts, and Exergalmics is an example of a reference that supplies the idea that when you're correcting a word, you might want to see the word where you're typing it before you correct it or complete it. So, strong case for obviousness. And on the secondary indicia, Your Honor, again, no strong evidence of commercial success. It's undisputed that Apple never practiced Claim 18 of the 172. So the idea that it was important to Apple's commercial success is not in this record. And there is no proof, again, of a nexus between commercial success and Claim 18. There was only evidence about commercial success and popularity of the iPhone generally. So, with respect, we don't see any case at all for secondary indicia overwhelming the strong obviousness case on the 172, and we think both 721 and 172 should be reversed on the grounds of invalidity for obviousness. Your Honor, there's nothing else? No, there is something. Ongoing royalties. Thank you, Your Honor. Ongoing royalties. Yes, Your Honor. As I understand it, there has been no permanent injunction entered yet by the district court. That's correct, Your Honor. So, but once that injunction is entered, then the ongoing royalties issue becomes moot thereafter, correct? Correct, Your Honor. Okay. But it's not entirely moot because even if an injunction is entered, there's still the issue of ongoing royalties from the date of the original judgment up until the date of the injunction, correct? That would be correct, Your Honor. But that could be handled also by supplemental damages, which haven't been resolved either. So, I guess, here's what, I'd like to take one step back. We think that it was incorrect to issue the ongoing royalties at the point the district court did because the district court lacked jurisdiction once Apple sought a review of the denial of the injunction. We think, second, they waived their right to ongoing royalties. And just to remind, there's no claim for ongoing royalties in the complaint. There's no request for ongoing royalties. I don't understand your suggestion that this could be handled as a result of supplemental damages because your arguments are that the damages have to stop with the date of the judgment, correct? Yes, Your Honor. So, Your Honor, it's certainly the case you must vacate the ongoing royalties. The question is do you remand or not to allow Apple to pursue ongoing royalties. And with respect to that question, I'm asking you to not remand to pursue ongoing royalties because Apple has waived and the court abused its discretion by holding that it did not waive because Apple never asks for ongoing royalties even in its injunction motion in May. It's not until after the injunction is denied in August that in September it files for the first time its request for ongoing royalties after it's already appealed the denial of the injunctions of this court. So, Your Honor, we think that Apple shouldn't get a second bite at the ongoing royalties questions when they waived it. We shouldn't cry a tear for them not seeking ongoing royalties. I understand what you're arguing. Thank you, Your Honor. What's left pending before the court with respect to damages? Just the ongoing royalties? Well, Your Honor, there's a number of things still pending before the district court. So, first, I don't want this court reverse the denial of the injunction and remand it for further proceedings on the injunction. We don't concede that any injunction should issue. But that remains to be determined. What also remains is when this court reversed the trade dress judgment on certain products. Those products were also subject to design patent and utility patent damages. And there's a scheduled trial, retrial on damages under the alternative theories for those products. And then on this is, I'm sorry, I've completely gone off into another case. I've been here so many times that I have confused two issues. So let me just consult for a moment to answer your question. All right. So, Your Honor, forgive me for misspeaking. There's nothing left to do other than to decide the injunction. And if you were to vacate the ongoing royalty order, we respectfully request that you not remand for ongoing royalties determination because it was waived. So that's a much simpler answer to your question. All right, Your Honor. And they were waived because Apple sought the injunction? Well, waived because they never sought ongoing royalties prior to... Ongoing royalties are a species of equitable relief. They sought an injunction. They didn't seek ongoing royalties after the trial when they sought the injunction. And when they appealed the injunction to this court, the district court had no jurisdiction. And by waiving their right to ongoing royalties, they can't come now and say, oh, we put it in the complaint by our boilerplate damages clause because this court has said in case that damages is... ongoing royalties are a species of equitable relief and not damages. Thank you. Will we store four minutes of rebuttal? Thank you, Your Honor. And why don't we add ten minutes to Mr. Lee's time to try to keep it even, but don't feel compelled to. All right. May it please the Court. My name is Bill Lee, and together with my colleague, Andrew Danforth, I represent Apple. I will try to address each of the issues that the panel raised during the course of Ms. Sullivan's argument. Let me just start quickly where the argument ended on this ongoing royalty issue. If you consider what Samsung is asking you to do, it's saying that for the period after judgment, through the date of entry of a permanent injunction, there would be no... It would be clear that it becomes moot after an injunction issue. If an injunction issues, it would become moot. So what we're talking about, Your Honor... Between the date of the judgment and the date of the injunction. Right. And the request that you reverse and find waiver is a request that for that period of time, for adjudicative infringement, there would be absolutely no relief. And nothing in this court's law nor in the district court's review of the relief that had been requested by Apple would suggest that's the appropriate determination. The district court specifically found there was no waiver, that there was a request for both equitable and monetary relief, and that's, in fact, what occurred. We did request injunction first. When the injunction was denied, we went back at that point in time and said, well, if we don't get an injunction, we should get ongoing royalties for that period of time. It is, Your Honor, just as you said, just this period of time, assuming that the injunction issues. The district court found it not to be waived. The district court had jurisdiction to address it because this doctrine of pendant appellate jurisdiction is quite... Why don't you turn to some of the substantive issues? Fair enough. So let's go to the... Can I start the 647 patent? And let me just say this at the outset. The issue which we're focused on today is a substantial evidence issue. To go to Judge Rainey's question, while there was no construction of these terms during the course of the trial before the last day, once you rendered your Motorola decision, the jury was twice instructed on the meaning of analyzer server and linking action. It was actually given the instruction in writing before the evidence was resumed, and it was charged at the end. So the question is, was there substantial evidence to support the jury's determination? And let me take them in the order... To some extent, it depends on what's meant by separateness. Because your expert admitted it's not... The library thing is not a standalone program. It's incorporated into the client program, right? No. In part, yes, and in part, no. He admitted that it wasn't standalone in the sense it performed... It operates in a vacuum. He never said it was part of the client program. And, in fact, a portion of the record that... It doesn't start out as part of the client program. In the sense, the client program is run, and the library program is called on when it's needed. So it's not run as a routine matter as part of the client program, but when it is called up and run, it is run then as part of the client program. No, Your Honor, that's not what the substantial evidence was. And let me answer your question directly and then talk about what the evidence was. There's no requirement that it be standalone. The requirement is that it be a separate routine. And there's nothing in the specification, nor the construction... Standalone would mean running separately. It's a routine that is separate, but routines don't run... No, no, but just to understand what you're saying, what you're saying is it doesn't have to be a standalone program in the sense of being run separately from the client program. Your Honor, again, I'm going to break it down. It has to be a routine that is separate from the client, and that's what the claim construction is. We agree, actually, that it has to be run separately from the client. It's a routine that runs separately, and that's what Dr. Mowery said. He said these three things. Where did he say that? He said it... Let me take you back to his testimony. He testified that the shared libraries are separate at 13035, and he said they were developed independently, that they're stored in different parts of memory, that there is only one copy of the shared library code, and the applications go to that code when they want to use it. That's at A13035. And to quote him, he says, and they use it there. That's specifically what he said to address this court's claim interpretation. He also testified that... Did he say it was run separately from the client program? I thought he said it was run separately from. Yes, and he said it at A13036, and the quote was definitely separate. And, Your Honor... What line is that on? Let me get to 13036. Sorry. So it's definitely separate from the application? Yes, line 20. Look at line 16 and 17 at that same page where you're there. It's a different part of the address space. It's in a particular part of memory, and all those applications go to the shared library code in one place that exists in the computer memory hardware to use it. So it's definitely separate. Yes, stored separately and not run routinely as part of the client program. But it seems to me that it is not run separately, and if separate here means run separately, it doesn't satisfy that. Well, actually, Your Honor, if you go to the next page, 13037, you will see that he says that when they want to use the code, which is what we looked at at 13035, they use it there. What line are we talking about? Let me get you the exact line, line 5. Wait, am I on 35 or 37? I'm sorry. 35 is what tells you that it's going there. 37, line 4, is the answer to the question. Well, it has access to the code, and it goes to the code where it is, and it uses it there. And it does that each time that it accesses the code. Now, Your Honor, without a doubt, their expert tried to make the contrary interpretation of the code. But to put this in context, there was, during the trial before the Motorola decision, testimony about why this claim was infringed. And it's important to remember that for nine of the 11 elements, there was no change resulting from the Motorola claim interpretation. Dr. Mowrey took the devices. He showed the jury how they worked. He took them through the source code. He then, when the claim interpretation came down that Your Honor referred to, he came back and said, same source code, here are the routines that actually perform these functions. And I'm going to get the linking function, because Samsung has completely confused two different concepts that have nothing to do with what Dr. Mowrey said. This is complicated. And so let me finish on the separate one. This is the circumstance where, on a substantial evidence question, there's nothing in the claim interpretation that says it has to be stand-alone. It says it has to be a separate routine. The question is whether it has to be a stand-alone program or whether it can be run as part of the client program, I guess. I actually think, Your Honor, two things. The question is, is it a separate routine, which is what this court's claim interpretation required. The jury was instructed that it needed to be a separate routine. Dr. Mowrey testified that it was separate in its development, separate in its location, separate in its execution. Now, to be sure, there are experts in this. I don't see the difference. 13035 and 13037. Where? 13037 says, well, it, the client, has access to the code, and it goes to the code, which is in the shared analyzer server, where it is, and it uses it there. And it does that each time that it accesses the code. And, Your Honor, unlike Samsung's effort, Dr. Mowrey gave an explanation for why in a smartphone, where space is the premium, right, and space for code is at a premium, that you don't want every application or every client to have its own copy of the code. It is actually working exactly the opposite. I understand that the copy doesn't copy, or at least there's testimony that it doesn't copy. There's also testimony that it's not run as a stand-alone program. That's the problem. Your Honor, there is no requirement in the claim interpretation that it run as a stand-alone program. And if the court were to consider... That's the question. Well, if you look at the claim, you will see that there's a reference to routines, and then analyzer server is one of many routines. If you look at the paragraph immediately preceding analyzer server and linking action, it says, memory storing information including program routines. The idea that each one of those program routines would have to run as a stand-alone is inconsistent with what's in the specification. The very program routines, and there are many in the specification, don't run independently in the abstract. They run as part of a series of functions. And in this particular circumstance, what happens is the client receives some data. And its question is, to reduce it to what's going on at least on my level, what is this data? So what does it do? It calls upon the analyzer server and says, help me figure out what this data is. The analyzer server then takes its code. There's only one copy. There is never a copy, as Dr. Mowrey said, back to the client. It runs its code, determines what it is, and then you get to the linking action. Then why don't you turn to the linking question. Here's the bottom line. The linking action has nothing to do with the glue code. And it's a complete obfuscation of what the issues were before the jury. The glue code is what allows the browser or the messenger to say, I've got some data. I don't know what it is. Here it is. Help me with it. The linking action is the specified connection. The specified connection occurs after the analyzer server code has done its work, in part. And it basically says, we've identified what the data is. And it becomes what's called an intent object. And Dr. Mowrey described this in detail. Once the intent object identifies the class of data and the specific data, so the class could be your telephone number and the specific data would be the telephone number, it then connects to something called start activity. That is specified connection. That's got nothing to do with the glue code at all. And you have to create a specified connection that then causes the CPU to do something. And that's exactly what happens. And Dr. Mowrey took the jury through exactly the process I've just described, to your honor, that basically demonstrates how the separate analyzer server function performs, how a specified connection is created as a result of that, and how that actually then detects the data and causes the CPU to perform the function that the user has chosen. It sounds like you're repeating an argument you've made before that these claims require only an association, associating, and not linking. No, your honor. I think, without a doubt, there was some debate about whether the claims were broader or narrow before your interpretation. We're not making that argument at all. There's a specified connection, in fact. If you go to 13039 to 13041, you'll find Dr. Mowrey's testimony specifically. And the specified connection is at A50868. It's the demonstrative that he used it. So what happens, your honor, is once the intent object is called and the data is identified and the class of data is identified, there is a specific connection made to start activity. And the reason you know it's more than an association is depending upon what the data is, class and data, the specified start activity will perform a different function. So if it's a telephone number and you want to call, start activity will cause the CPU to call. If it's a telephone number and you want to save Mr. Danforth's telephone number to contacts, it will cause the CPU to do something different. Without a doubt, the claim interpretation that this court adopted is narrower, but you can satisfy the general and the specific, and that's exactly what happened here. In fact, at the trial, and Ms. Sullivan mentioned this, suggesting that Dr. Mowrey had changed his position. Actually, the one expert who was admonished on the last day of trial for suggesting that somehow his positions were consistent with what this court had done was their expert, not Dr. Mowrey. Dr. Mowrey took your claim interpretation and he walked through it and said, yes, it satisfies the general, but it also satisfies the specific. And at the end of the day, the question is, I think there are two going to judge your Honor's questions, which are does the claim require a completely stand-alone separate routine? And this court's Motorola interpretation doesn't require it, the claim doesn't require it, but it does require a separate routine. Dr. Mowrey said the routine is separate, that there's one copy of the code, it's called when the client wants to use it, it's used at the analyzer server, and once it's used, there is the creation of a specified action, which is the linking action. Last comment, Your Honor, to your question is the claim term is creating a specified connection, and when the intent object is identified and the connection is made to start activity, a specified connection is created, so it specifically satisfies claim limitation. For all those reasons, and I'm not suggesting that their expert didn't say something different, but the jury heard both experts twice. The jury saw the devices, the jury had the source code, the jury had the routines, and critically important, it had the explanation from Apple's expert as to why he was correct, that there's only one copy of the code and it's used at the analyzer server. And it's because it would defeat, to have multiple copies of the code would require the use of space that's at a premium, an absolute premium in what is basically a small... That goes to a, I guess a confusion I have with respect to separate from. You're saying that this does not have to be standalone, be it, it's clear it's got to be separate from the, what you've identified as the client, the browser and the messaging applications, that's the client. But all you've argued before us is that this is in a different location, it's got a different address, and I'm not, to me, I'm not convinced yet that that's separate from the fact that you have a different address or a different location, and you even take your argument further and say we can't have many locations, we've got to have at least one location because of, you know, space is precious here. I'm still not totally convinced that this is separate from. Your Honor, here's, I can answer it in two parts. Okay. The reason that it's not standalone, as Judge Dykes used the term, is the client... And your expert used the term. Yes, and for a good reason, because it is the client who has to say, I have this data, what is it? Help me determine what it is and help me do a linking action. So, by definition, the process has to start somewhere. But it's not sending a query to the library program, it's actually incorporating the library program and running it. Absolutely not. That, you've just articulated what Samsung's expert said, a theory that was rejected by the jury. What our expert testified to and the jury accepted and the district court credited on JML was, it sends the data that it has seen, but doesn't know what it is, to the separate analyzer server, which is physically separate, independently developed, different location, one copy of the code, and it operates on that data there to figure out what it is. How is that consistent with it not being a standalone program? Well, Your Honor, there's no requirement legally that it be a standalone program. What you describe, it seems to me, is a description of something that's a standalone program, whereas your expert said it wasn't. No, Your Honor. What I've described is a standalone routine. And if you look at the claim... Routine as opposed to a standalone program? A routine that operates separately. And that's exactly what's going on here. If you would be taking the claim interpretation that was given in the Motorola case and reinterpreting it to require something more specific, and that wouldn't actually make any sense in the context of the specification, and it certainly wouldn't make any sense in the context of the evidence. No one's suggesting that the analyzer server can start operating without something happening at the client. The big fight was this. This was the big fight. Does the code from the analyzer server get copied and become part of the client where it's run? Or is there one copy of the code and is it run at the analyzer server separately as a separate routine? Well, I don't think being... It's certainly not copied. At least you have testimony that it's not copied into the client program. Well, we have our testimony saying it's not theirs to say it is. No, no, but that's not the issue. It's not whether it was copied into it. The question is whether it's run separately. And Dr. Mowrey said it was. And the debate that you and I are having here, or the discussion you and I are having here, is the disagreement the two experts had. So we have one of two issues. Either the court thinks that there is appropriate to have a new claim construction that's actually more specific and different than what the jury charged. No one has urged you to do that. There is no suggesting the claim interpretation is wrong. The only question is, on the claim interpretation, was there a substantial evidence? The question is, what does the claim interpretation mean? No, Your Honor, they haven't raised that issue. There is no issue of claim interpretation. There is no issue of whether it was incorrect. There is no issue of whether it was given. The only question is, on that charge, was there substantial evidence to support the jury's decision? So let me quickly try to address some of the other issues. Yeah, I'll turn to the 721-172. Now, let's hypothetically assume that, with respect to each of these, we decide that this is a strong prima facie case of obviousness, and so why did the secondary consideration from each instance overcome that? Your Honor, because in each instance, the combination of – and I'd like to get back to the prima facie case, but I'll take your question as it stands. And, again, while obviousness is a question of laws, the court's articulated it, these subsidiary or these underlying facts are judged by substantial evidence. And in both circumstances, there was industry praise for the feature. There was demonstrated commercial success. There was the conjoint study. How was there demonstrated commercial success with respect to these features as opposed to the iPhone as a whole? Your Honor, because the testimony went to these features. The conjoint study went to these features. The survey results went to these features. The copying evidence, we're not talking about – and I don't think there's a distinction between copying the feature and copying the words of the patent. The features are going to come out almost by definition before the patent emerges. Copying a feature, I mean, would you have – I understand, but what about – what's the evidence of commercial success? The conjoint survey? The conjoint survey is, in part, the evidence of commercial success. Is there a link? Was there evidence linking the survey to the specific features on the patents? Yes. For the 172 patent, there was, at 51440, you will find the chart that Dr. Hauser used that gave a specific indication linking the – giving the nexus between the claimed feature and the success of the product. There was a comment about whether we practiced precisely the claims that were asserted. I think the court knows that there was a limitation on the number of claims that could be adjudicated to 10, and that included the ones that you could claim you practiced. So in one case, while we didn't practice the precise claim, we practiced another claim of the patent, but it was just not something that could be put before the jury. But in this case, and the court has addressed it in part in the permanent injunction opinion, there was industry praise for the specific feature. Can I ask you just a more background question? Sure. With respect to both patents, you make an argument. You think it's strong. It may or may not be strong, which was the 721 teaching away. And so you make a strong argument, and the case was made, that there was no really strong prima facie case about it. Right. And the JMAL, the district court, bought that. And then both you and the district court, in your brief, you say, and secondary considerations add to this or bolster this or whatever. No, I don't read anyone that's ever made the argument that if we rejected your case, your arguments with respect to teaching away, for example, in the 721, that you are arguing that secondary considerations by themselves, in the face of a strong prima facie case of obviousness, would be sufficient to dislodge. So let me answer in two ways. There is a disagreement in some of the court's opinions about whether you view it as a prima facie case and then move to secondary considerations or you view it as all this evidence collectively to determine whether it's obvious or not. Viewed at the end of the day, obviously, you have to consider all of it. Right. And I think my answer is twofold, but I'm going to take your question first and then answer in a way at least that I sort of understood the issue and as we've raised the issue. Have we made the specific argument that if you adopted a prima facie case argument that the secondary consideration would overcome the prima facie case? I don't think we said in our brief. I will say to you here that I think that the answer is it does, and the jury was entitled to conclude that it did. I mean, these are circumstances not where there is some vague accusation of copying. This is the situation where the jury had before it and you had before it literally a copy of our phone and our feature, literally a copy of the prototype that Samsung had, literally a label of the feature, and literally a direction at the bottom to copy our feature. So you have industry praise for autocorrect and slide unlock. You have the kind of studies conducted by Dr. Houser, but from Samsung's own mouth you have them saying this is a great feature. Our customers, the carriers say that our version is jarring. Here's what we're going to do. And they adopted the feature which was sound to in print. There's nothing in this court's law that says in order to there be evidence of copying, you have to show that someone read Claim 18 and then religiously followed Claim 18. This is a circumstance where our product hit the market. They identified it. They identified these features, autocorrect and slide unlock, as important, important to them, important to their customers, and then they religiously copied it. Wait, where do they say that, that it's important to their customers? Your Honor, there's a specific, I don't know if I have this specific page for you, but there is, I'll get it for you, but there's a reference to the jarring aspect. I know that. I've seen that. That comes from the customers. Yeah, but I don't remember any testimony or any document that said it was important to their customers. Well, if you have a document from Samsung. But you're not claiming that there's a specific reference to it being important to their customers. No, I think what I intended to say is you have a document that has… You're inferring that from the jarring. Yeah, it has ours on the left, theirs on the right, and a statement that the carriers say that their feature is jarring. And then the recommendation is to adopt our feature. That's pretty strong evidence. And so I think the answer, Chief Judge Post, to your question is we would say that it does. Now, very quickly, so I also can get to both the heuristic, to the heuristic issue. For both the 172 and the 721, we think the evidence before the jury supports the non-obviousness determination. And just to summarize what your Honor said, in one case there is placent, which specifically refers… It's for a large, an air conditioning unit, it's a wall-mounted unit. It refers to the difficulty of implementing it, a complex slide on the left. That is teaching away. Right now, there was disagreement, but this court's spetrolytics decision said… I don't understand why there's teaching away. I mean, you had the other reference that showed, on a phone, sweeping from left to right. And what was missing from that was a depiction of an arrow or whatever it is moving along with the finger that's doing the sliding. That's what's missing from that. And looking to placent, that's where the visual depiction comes from, right? So where's the teaching away? Right, Your Honor. Placent says that at 20743, that sliders are not preferred. Sliders are more complex. Sliders are more difficult. And that's followed by quite a lengthy statement that says, even if sliders are not preferred, the fact that users… And then it goes on and on to talk about the advantages of sliders. For sure, Your Honor. And if this court held in spectrolytics, when you have… And I think Judge Koh described it as there was interpretations on both sides. The jury heard testimony from both experts as to how one of the ordinary skill in the art would interpret this reference. Why does something teach away if a prior art reference says this is preferred, but you could do it the other way and there are advantages to doing it the other way? That doesn't seem to me to be a traditional teaching away. Well, Your Honor, I think the best I can say is this. It was a jury question of whether it teaches away or not. You had two experts who got on the stand, looked at the language, that sliders were not preferred is what it says. Not that they are preferred. Were not preferred. Sliders are more complex. What case says that if something is not preferred, but it has advantages and could be used, that that involves a teaching away? Your Honor, I don't have a precise case that says precisely that. But what I do have is spectrolytics that says that the question of teaching away is a jury question. The jury, everything that we have been discussing back and forth, the jury heard from both experts and they heard from us in closing argument. My review of Poisson was that sliders were not preferred, but still the authors, they saw some value in them. For sure. They evaluated them. Not teaching away. Now the question is, if you have Poisson, is there the motivation to combine? Just to remind the court, when Judge Coe considered this preliminary injunction, she said it's not enough to say you just would combine them. We got to trial and all there was was, well, it would be natural to do this. But why? No explanation. This is a situation where the one thing, if we were discussing this in the abstract in the first instance, but this is a situation where the jury heard about Poisson. The jury heard about Neonode, which actually is missing more, I think, than is described by Samson because without the unlocked image, you can't do three parts of the claim. Then the question becomes this. Given that the jury heard from both experts about Neonode, given that they heard about Poisson, given they heard, the only suggestion of a motivation to combine would be that it would be natural to do so but nothing more specific, and given that they heard all the secondary And the answer is yes. Can you move to the questions I asked regarding your answer? Sure. I think I can answer them quickly, Your Honor. If the court considers the district court opinion I'd like to point to the evidence that shows some limiting factor or some sort of narrowing of the definition of what a heuristic is other than the fact that it's a good idea or it's a set of rules or a set of principles. I can give you three buckets of evidence. The first bucket is at H77F2 at H65-66, which is also referred to by the district court at A181. This is where the term heuristic was interpreted on preliminary injunction from the related 604 patent. And the court referred to these three pieces of evidence. And then I've got the two other buckets. First was the party's agreement that a heuristic was a rule of thumb and that it had a meaning to a computer scientist. The second was dictionary definitions from the field defining a heuristic. And third, statements from the file history of the 604 patent distinguishing a heuristic from a constraint satisfaction parameter. She had all of that before her because she didn't terminate it. She had determined the meaning of heuristic before. At trial... So a heuristic is any rule except a constraint satisfaction parameter? No, as I said, it's a rule of thumb. It's not a constraint satisfaction parameter. And the example that was given to the court... So that means it's any rule that's not a constraint satisfaction parameter. It's a rule of thumb, and there was an example... What does the thumb add to it? Well, actually, what both experts said... And if I take you to A50095... At A50095, Samsung's expert described heuristics as rules of thumb that were well recognized in the computer industry, generally implemented in the past by computer scientists, and, to quote him, generally recognized as rules of thumb by computer scientists. That could be anything, correct? No, Your Honor, there actually was an example, and if I take you to A50894, you will see the difference between a heuristic and a constraint satisfaction parameter. It could be anything except a constraint satisfaction parameter. It's a rule of thumb to be implemented by the computer program. It's not a constraint satisfaction parameter, yes. And the question is, does one of ordinary skill in the art, as opposed to you or me, recognize it? And two critically important facts. They're asking you to find a term indefinite when their expert said it's well recognized in the field, well understood in the field, and implemented in the field. How could that make common sense? What's the third bucket? The third bucket is the fact that both experts applied the term heuristic, both to the prior art and to the infringing devices, and were able to do so with reasonable certainty. And the citation term... You've given me just a very broad definition. I mean, a rule of thumb is pretty broad. A rule of thumb recognized by experts as having general application is broad, and in my view, indefinite. What evidence is there in the record that would take me to look at that and say, okay, here's evidence introduced that narrows the definition or applies a definition of what a heuristic is in this particular pattern? And your Honor, let me give you... The question is, would one of ordinary skill in the art understand it? Let me give you more fully the quote of 50095 by Foote. Because, Your Honor, the task on indefinite is not to identify every single... I understand what you're saying about one skill in the art, but what skill in the art has got to base a definition or their view on the pattern? So, where in the pattern? Where do we find that one skill in the art would look at and say, okay, I understand what the heuristic is in this case. The best I can give you is the following. One is the testimony of their expert at 50095 that heuristics are... 50095? 50095. Yeah, clear. Right. Paragraph 291. Paragraph 291. Paragraph 291. This is where he's allowing... trying to allow himself some room to maneuver. And so he says the second sentence, while the term heuristic is ambiguous even under this court's construction, there are techniques that have been recognized, are generally recognized, which would be generally... or would generally be recognized by a computer scientist, and one of skill in the art as rules of thumb that do not consist solely of constraint satisfaction parameters and therefore are unambiguously heuristics. Okay, so the term techniques. There are techniques that have been recognized. What are the techniques in this pattern that heuristic is referring to? Well, the techniques, the specific techniques that are described in this pattern are the ones that the district court referred to in his decision at 877F sub 2nd, where she went through and talked about how the term was used in the patent, how it had been used in the file history, how it was distinguished from a constraint satisfaction parameter. That's just saying that it's anything, any technique, other than a constraint satisfaction parameter, right? It's not, Your Honor, because we're not... This is not respectfully an inquiry directed to a lawyer. It's directed to one of ordinary skill in the art. And if a heuristic has meaning to one of ordinary skill in the art, as their experts said it did, as our experts said it did, then that's sufficiently definite. We ought to focus on what Samsung's argument is. Their argument is that you have to be able to identify everything that falls within the genus of heuristics. And no case in this court has ever suggested that that's a requirement. You have a circumstance where our expert says that it has an accepted definition. Theirs says, to quote him, are therefore unambiguously heuristics. And then both of them applied the term heuristic, a rule of thumb, to the prior art, the freeway system, or its disused product. And they had no trouble doing that at all. Okay. Thank you. Ms. Sullivan, to keep it even now, we'll give you 12 minutes of rebuttal, only if you need it, just as we went over 14 minutes on the other side. Thank you, Your Honor. I'd like to start with the 647 patent. Can I just get one further clarification, just to make sure there's no misunderstanding? I understood you to tell us that on the 959 patent, if we were to affirm the J. Mulls non-infringement, you do not see that we need or should pursue the question of indefiniteness. That is correct, Your Honor. We think you can affirm the judgment of non-infringement and not reach the indefiniteness question. If you do reach it, we think it's indefinite, for reasons that Judge Randy's questions have already explored. On 647, I want to start with analyzer server being separate, because as Judge Dyke pointed out, that is the key question. And I was astonished to hear Mr. Lee say that analyzer server is operating separately when running or executing the detecting and linking functions, because there is no evidence of that. The key question is not whether the shared library code is separate when it's stored. It's not whether it's separate when it's placed in memory. The question is, is analyzer server separate when it is detecting and linking? That's what this court's claim construction was in Motorola. And there is no evidence that it's separate when it's doing the detecting and linking. And Dr. Mowrey's important statement at 13036 and 1307 does not say that the library is acting separately. So my friend, Mr. Lee, pointed you, Your Honors, to the top of 13037 and to the sentence from Dr. Mowrey, it has access to the code and uses it there. But the it in that sentence is the messenger and browser application. If the application is using the shared library code, the shared library code is not being a separate analyzer server and doing the detecting and linking. And there was a dispute about whether the shared library code is copied into the application. Your Honor, that's beside the point. The dispute about whether it's copied, there was no dispute that the code goes in Mowrey's own words, the application goes to the shared library and imports or incorporates it into the code. The commands in the source code say import shared library code. That's how they read in the source code. Your Honor, the lack of a stand-alone analyzer server function is crucial here. There is no evidence that the analyzer server or the shared library code is doing any detecting independently on its own. And that's your construction. Your Honor was quite right that the testimony by Dr. Mowrey about the shared library code being separate was not about it being separate when it is running. It can't run on its own because it can't run as a stand-alone. That's Mowrey's own admission. It cannot run as a stand-alone code. And because it's not running alone, there's no separate analyzer server. And that's enough to decide the case. But I gave you a second dispositive ground for why we went on 647. And I'm afraid Mr. Lee created quite a bit of confusion about it. So I want to go back and see if I can say it one more time very simply. This is my glue code argument. And it has nothing to do with specified connection. Your construction gave us two important readings of the claim. Analyzer server has to perform linking separately from the application. And that's what I was talking about. You also said that the linked action has to be through a specified connection. But I wasn't talking about that when I talked to you about glue code. What I'm asking you to do is go back and look at Mowrey pre-Motorola where he says there are three routines that do the linking. And I want you to go back and look at Mowrey post-Motorola and see that he points to the exact same three routines that do the linking. And this really takes only six pages of the record. You just look at 10.857, 10.859, and 10.863. And you'll see the three routines that do the linking according to Mowrey. Then you go to 13.033, 13.034 after the Motorola construction. And you're going to see those same three routines that do the linking. This has nothing to do with shared libraries. These routines are in the application. And the key point is when you go back to his post-Motorola construction and you ask where's the code that does the linking, he says, again, it's in the client or from the client. It's in Messenger or from browser. And I gave you the lines on that. It's 13.033 lines 8 and 20 and 13.034 lines 19. So, Your Honor, we think we win on shared library because they're not operating standalone. But we think we win even without getting to shared libraries because the linking code is in the client. And if anything that's in the client cannot be an analyzer server or a server routine because it is not separate from the client. I'm not arguing anything about what our expert said. Mr. Lee tried to put this back into a battle of the experts mode. I'm relying solely on Dr. Mowrey, Apple's expert in his admissions. And it's all there. There's no separate analyzer server that is performing the linking. And there is, second, no separate analyzer server that's performing the detecting because the application is importing or incorporating the library code into it in order to do the detecting. Analyzer server is not met and the 647 judgment should be reversed. Now, I'm sorry to belabor it, but I just thought it was very important to clarify that we win on analyzer server without getting to specified connection. Your Honor, Judge Reina, with respect to your questions about specified connection, I couldn't agree more with you that what Apple is doing here is going back to its pre-Motorola construction of associating. But I'll leave to the brief the explanation of why. We think linking means adjoining. That's what you said in Motorola. What you need to join is the user selection of an action and the performance of the action. In the old world, remember this patent is a very old patent. It arose not in a world of mobile phones. It arose in a world of desktop computers. In the world of mobile phones with Android and an open platform, you don't have a linking and joining between the user selecting an action and the call being made or the phone number being stored. It has to go through that whole long process that Mr. Lee told you about, about intent objects, populated start activity, which then asks for a world of applications that might perform the call or store the address. That is not linking or joining. So we think we went on specified connection, too, but I'll leave it to the brief to explain. Your Honor, I'm 721. Did I understand Mr. Lee to say that this court's, let's say, interpretation of penalizer's sermon was read to the jury? Yes, Your Honor. He's correct that we're not disputing the instruction, but we are strenuously disputing that any reasonable juror could find infringement under this court's instruction, construction as instructed to the jury. So there's no substantial evidence to support it. So that's correct. Your Honor, on 721, I want to correct a couple of factual points. First, the conjoint study does not support commercial success for the 721. Mr. Lee cited you to 51440, but that's Hauser on the 172 patent. Hauser did not test the 721 patent for phones. Second, on teaching away, there's no factual dispute on teaching away. Sorry, there's no factual dispute on what the reference has said. And the dispute on teaching away, we would respectfully suggest, is a legal question. Your Honor, I think the answer to the question, does Plaisant teach away, is a legal question you can answer. You looked at the page, you cited the page where Plaisant talks about how the use of sliders is encouraging, and the use of sliders is an effective way of, it says positive things about how it solves the inadvertent activation problem. And as a matter of law, it can't be the case that a reference that says that something has advantages, but, you know, it talks about something is preferred and has advantages, but it's not always preferred because there might be other ways of doing it. There might be two things the reference points to. That's not teaching away. And, in fact, this court in Allergen and Goderma, which we cite in our briefs, has pointed out, as a matter of law, that there's no teaching away. So if we have weak evidence of secondary indicia, we have strong prima facie case of obviousness, and we have a legal error on teaching away, we respectfully suggest that you should hold, as a matter of law, on the 721, that it's invalid as obvious. On heuristics, Your Honor, we don't think you need to reach it, but if you reach it just because an expert can construe the claim does not satisfy Nautilus. It has to have boundaries. And I didn't hear any boundaries in Mr. Lee's account. And I think I have nothing more to say, Your Honor. The key thing is we ask that you reverse on 647 because under your claim construction, you can't find infringement. No separateness between the analyzer server or the server routines and their performance of detecting and linking functions. And please focus on the language of the claim. It's not an analyzer server in the abstract or a routine in the abstract. It has to be the analyzer server or server routine for detecting and linking. So the key to the separateness question is to ask, is there separateness when the detecting and linking is being done? And there is not. Thank you, Your Honor. Thank you.