Oral argument — Bright Data Ltd. v. Code200, UAB

The next case is Bright Data v. Code200, 23-2144. When you're ready, Mr. Harkins. Thank you. May it please the Court, Bright Data filed this appeal to correct incorrect claim constructions that were applied by the Patent Trial Appeal Board and the Board's constructions of the term client device and server unfortunately ignored the very problem and solution that was presented in the challenged patents in this case and described in column 2. We have one specification across all the patents, it's the same. The terms in the claims are the same. And the patents describe the problem in column 2. The problem is that proxies or proxy servers that are talked about in column 2, line 10, I'm using the 342 patent for all my citations because that was the first of the IPRs, and so this is appendix 11-59, talks about this use of servers. And it talks about them requesting data from web servers on behalf of clients. But there's a problem. And the problem that's identified is that servers are too expensive and they're not practical to locate everywhere that you want them around the world so that they will be near the content that you're trying to obtain. So, Counsel, appellees point to several instances in the spec that seem to suggest that a single device can serve different roles depending on the particular connection. How does your construction account for those portions of the specification? Well, that's the solution, Your Honor. So, Judge Kennedy, I'm going to point this out, and I think it's exactly right. The way the patents, so a server is a commercial device. It is something you could, at the time, a person who scored it in their skill could go to Fry's or some other place and go say, I want to buy a server, and they would give you hardware. And that hardware was intended to be in a network and always on. That's what a server was. And client devices, which generally PCs or laptops or could be even a phone, that were not intended to be in the middle of a corporate network and that weren't always on. Those were the devices. That's what the devices meant. But there's also the hypertext transfer protocol. And one example of that that's talked about in the patent is RFC 2616. And that talks about something called the client-server model, which talks about the roles that different kinds of computers play. And so if a computer is asking for... So maybe can we go to column four, lines 41 to 53? Yes. Just to kind of ground our discussion that we're having here. When you get to that section... Are you there? I am. Okay. And I'm in the 342, appendix page 1160. When you get to that section, you can see a discussion about each communication device may serve as a client, peer or agent. Right. Right? And to me, at least that portion is supportive of having constructions that are kind of role-based as opposed to hardware-based, which I think is what you were starting to describe to me. Why do you say that... I'm assuming you're saying that that is the incorrect understanding of that portion of the specification. If you are saying that, I want to know why you think that. So the idea of a role that something plays is different from what they're talking about as the communication device itself, which is the hardware. So what they do is they take what's being referred to in the claim as a client device, which it says right in the patent, and one of our arguments is lexicography, is a computer of a consumer. This means... And this is not a commercial device. So the problem that's been identified is proxies are good, but they're expensive because they're in this commercial server hardware. And what we want to do is create a whole new network where instead of using commercial servers, we're going to use these communications devices, these client devices. And those devices are going to act as proxies instead. And the benefit of that... I'm still not certain you've answered my question. I think I understand what you're trying to describe to me, and I think you're very much making a hardware-based argument. Correct. What I was seeing here in terms of COM4 is that it could be more of a role-based. And what I want to know is why you think that is incorrect, assuming you do think that's incorrect. So the reason it's incorrect is because the whole point of this patent is to take one of these communications devices and use it in an unusual way that hasn't been done before. So it's actually taking on a different role. So these roles of client, peer, and agent, what it is is you start with the client device. You start with the hardware that is a PC or a laptop or something out in the world, and that's what's talked about in column 2. Column 2 actually defines this at line 45, saying, Computers of consumers herein referred to as client devices. So it literally is saying, I'm not using the server devices. You're saying that's lexicography. Isn't that an awfully subtle way to be a lexicographer in the background of the investigation? Background of the invention, excuse me. In column 2, in the middle of a paragraph, referring to a specific figure, which is admittedly prior art, that the inventor meant to carry that as a definition of client devices all the way through to the claims. Wouldn't that be an awfully subtle way to do it? What I will say about that is, when you say something is referred to as client devices blatantly with this type of language, referred to herein as, in the Kyocera Senko case, that language was pointed to. Why wouldn't one of the art reads as herein referred to just to be limited to the discussion of the prior art, given the placement of the statement and the specific reference to the prior art figure? There is no difference in the way client devices are talked about in the prior art, as far as devices go, versus the prior art versus the invention. It's consistent throughout the whole patent. There's just no... But to win the appeals, I understand your response to Judge Cunningham. You need us to say that client devices are a specific type of hardware, and it's not just any kind of hardware playing the role of client. Isn't that what you need us to say to win this appeal? That's correct, because that's the solution that the patent... The problem is they say there's a specific kind of hardware. So the only way you get there, I think, but tell me if you have some other way, is if we read this sentence as lexicography. Otherwise, you have no way to get where you need to go. Well, with all due respect, Your Honor, I think there's several ways you get there. Lexicography would answer the question. So if you agree with this on lexicography... What's the other way? The other ways would include that there is prosecution history that says that a client device is not a server, and a server is not a client. And that is something that... Can you take us to that particular JAP so we can discuss further the prosecution history that you want to rely on for that point? So in the area of Appendix 3907 through 3934... 3907? I'm going to need you to be more precise than that. So if we could turn to 3410, for example. 3410? Yes, please. Oh, my gosh. So here, in the discussion of 3410, in Argument No. 1, in the discussion of Garcia, it says there's a clear distinction in the art and is taught by Garcia reference between clients and servers. Client devices, such as Client 104 in the Garcia reference, are end units that request information from servers. And then this says, at the end of that paragraph, as shown in Figure 2, a client device typically connects to the Internet via an ISP using a single connection. That is like a PC type of operation, not a commercial server. Then we go to B, and it says, in contrast, server devices are known in the art to be dedicated devices to store information and objects to be provided to clients upon request. That is, these are not the kind of devices that go on and offline the way it talks about client devices and communication devices in the patent. They are dedicated devices that stay in the network all the time, which is one of the reasons that they're so expensive and they don't have the flexibility of being everywhere all over the world. That's the problem, and the solution was that you're going to use these client devices instead. So we're already distinguishing those things in 3410. Then at 3413, at the bottom of that page, it says, however, the Garcia reference in general and the cited paragraph in particular are silent regarding selecting non-cache server in general, a non-server in particular, and further regarding selecting a client device as recited in this claim. That is, the client device is a non-server. It is not a server. These arguments were being made in the original prosecution here. So that's how the claims were to get over the reference Garcia, which is not an issue here. There's a lot of prosecution history here. The red brief at 33 to 34 takes us to the 614 patent, column 119, it's at page 18748. Can I take a note on this? Yeah, 18748. It's the 614 patent, which I think is in the family here. Column 119, round lines 18 to 20. Alternatively or in addition, each of the devices denoted herein as servers may equally function as a client in the meaning of client slash server architecture. How does that fit into your argument? So the 614 is a different patent, different patent family. It doesn't claim priority back to these patents. It has a whole different specification. And in that patent, it talks about, in that system that's talked about in the other patent that's not at issue here today, that different devices can be used interchangeably for that invention that they're talking about. Basically, what they're saying there, and then in the figures of that patent, and I just don't have it pulled up in front of me, but in the figures of that patent, they actually have the idea of pairing a client with a server, and they do show in the figures that there would be different devices. But for the 614 patent, which is a later date, different priority and everything like that, they say you can, for our invention purposes, you can mix and match components to perform what you need to do. And the reason that can't be done in the patent we're reviewing is the prosecution history that you've pointed to. Well, that's a different point in time that's later talking about a different set of inventions. And the prosecution history here is saying there's a difference between client and servers. And even what Judge Cunningham had pointed out about the clients, peers, and agents, all of those roles, if you look at figure six in the patent, that's software that's loaded onto a personal computer. That's not put onto a server. That's the whole point of the patent, is we can't put servers everywhere we want them around the world. It's too expensive. Those are dedicated, always-on devices. So what we need to do is we're going to have this other software everybody already owns. Let me go to your, I think you have the baseball analogy. Is that right? A what? Baseball analogy? About catching, pitching, and catching. Yes, that was your analogy, correct? Right. And so my understanding of the way you think your analogy would work was if somebody's labeled a pitcher, they're always labeled a pitcher, and if somebody's labeled a catcher, it's always labeled a catcher, even if the pitcher's catching or the catcher's pitching. True. Okay. But it seems to me that that analogy actually makes more sense to support the opposing counsel side, right? Because when a person is pitching, even if they were previously a catcher, I would think they would then convert to being a pitcher and vice versa. Do you understand how I'm seeing the way I think that analogy would work? I think this comes back to this idea, which is there's a difference between what was talked about as roles in software. That is, do you have a client program running on a computer that asks for information and receives information? Or a program running on a computer that would be a client, that would act as a client? A program running on a server that receives requests and sends back responses? There were devices that existed in the prior art and are talked about in RFC 2616 known as proxies that do both. They act as a client and they act as a server. Those were only server hardware. Client devices never did both of those things. That's what there's a novelty in the patent about is that proxy, and that's what it talks about at column 2, line 10 and 11, were proxy servers. Isn't RFC 2616 also discussed in the actual patent specification, right? Correct. And so with that in mind, that it would be considered also part of the intrinsic record, right? Yes. And that RFC... I'm sorry, I'm messing up the letters. That RFC 2616 actually talks about, which I think is what you were just getting to, that a client device could act as a server and vice versa. Not client device. That is a program called a client program that has a certain functionality. It's not a device. It's just a functionality. Okay, so tell me what you think I should take from RFC 2616. That's what the roles are. That's when you have different roles that can be played. In the prior art, a server could act as a client and a server if it was a proxy server. But what we're talking about as a server is this commercial hardware that was always on. In the prior art, you never had a person's computer or phone or PC that is a client device acting in the role of a server and a client between two servers. That never happened before. The first time that ever happened was this patent. And the only way that you can get any meaning out of the problem and the solution here is to recognize that. That the whole point here was to... Okay, Mr. Harkin. We've asked you a lot of questions, so I'll restore your rebuttal time, but you're out for now. Oh, thank you. Thank you, Your Honor. Mr. Leventhal. May it please the Court, Daniel Leventhal for Apple East. Your Honors, what Phillips tells this Court is that the plain and ordinary meaning that the Court is seeking for a term is that which is consistent with the specification. Bright Data asks this Court to violate that mandate twice, first for client and second for server. Magistrate Judge Payne, Judge Gilstrap, eight panels of the PTAB, each directly considered this, the hardware-based construction of client and the hardware-based construction of server, and rejected them explicitly because they are not consistent with the specification. Judge Cunningham identified the portions of the specification to which I was going to point as showing it is consistent with the position, with the role-based construction, as does RFC 2616. So what I would like to just briefly address is a couple of concepts that my friend on the other side addressed in terms of the problem and solution. The problem and solution that was described is not the problem and solution that this patent actually addresses. The problem this patent addresses, and it begins in the abstract, is that getting content from a web server is bad. It's a congested part of the freeway that you want to avoid. That's what the abstract says. And the goal of this patent is to find a way to avoid going to the server at all costs. As a last resort, you'll go there, but the preferable thing to do is to grab a stored copy somewhere other than the web server. And in figures one and figures two, the two prior art embodiments, the inventors are describing two known ways to attempt to stop, to get around this, and explaining why they're no good. Figure one, disagree that figure one suggests anything about the structure of a server generally. What figure one says is, let's do a content delivery network, what Akamai has popularized today. And what they say is, if you did one of these content delivery networks and you wanted to store all of the stuff that was relevant, then it would be huge and expensive and everything else. It's not a definition of server hardware generally. It's saying, if you wanted to do that, they say it's a bad idea. And so they move on. Figure two, they move on and they say, well, what about peer-to-peer networks? That has a great idea because in peer-to-peer networks, the content is distributed. A little bit of content is stored in different places. Can you directly respond to the lexicography argument that the Housing Council makes? Yes, you are. I'm pointing to that here in language. Yes, you are. It is not lexicography. As Judge Stark explained, if you look at the beginning of the sentence, it states that what it is talking about is in the network 50. And, Your Honor, I'm at column two, line 45. And if you go up one more line, it says, figure two is this peer-to-peer prior art network that has file network, transfer network 50. And it states, Your Honor, that then the sentence, that's alleged lexicography, is explicitly limited to in network 50, which is the prior art network. And you go to the end, and there is an incomplete quotation in the brief. It does not say, here as referred to herein as client devices. It says, as referred to herein as client devices 60. Client devices 60 is a specific element only found in figure two. In figure three, which is the actual embodiment, everything in there is a communication device 200. Can you also respond to the prosecution history argument that opposing counsel made? So, number one, so the prosecution history, in referring to 3410, the response to that from the board, which is appendix 356, explains three points, all of which independently address that prosecution history. Number one is that this is a grandchild. This is a grandparent. It is not using any of the claim terms at issue. What is being described in Garcia when they talk about the client and the server is how a poseda would interpret and understand a specific word in the prior art, and therefore whether it would meet the claims. We think that's well short of clear and unambiguous. The point number two is the court cites to Epistar and says it must be an expression of manifest exclusion or restriction. And especially important, Your Honor, I believe is in part .3. And then again, this is appendix 356. They point out that the prosecution history comments cannot trump the plain language of the claims and the direct teaching of the specification. And they're citing to the Talcordia case. And the bottom line, Your Honor, is none of these are disclaimers. And by the way, the 936 patent, even though it is a grandparent of what we're talking about here, that file history excerpt was 2017. The priority date here is 2009. You cannot use file history statements to rewrite history and contradict the specification. And that is why we believe it is not a disclaimer. And that applies for all of their arguments. So I thought that I was onto something here with the, I'm gonna say RFC, give me the numbers again. 2616. 2616. But then when I heard opposing counsel discuss it, when I was trying to explain how I thought it supported up the role sort of based claim construction as opposed to the hardware based claim construction, they drew a distinction between client, program and client device. And I just wanted to hear you speak to that as well. Sure, yes, Your Honor. So number one, there is no such distinction if you go back to the specification, because I'll direct the court to column five, lines 21 through 34 of the specification. And in that sense, recall, we're talking about figure three. And in figure three, we have little circles that are labeled agent, peer and client. But what the specification explains is, even though they have those different labels, they're all communication device 200. They all have generic components of a computer. And by the way, just very briefly, we disagree that the specification ever suggests anything in figure three is a personal computer. It says they are generic components of a computer. And that's column five, lines 54 through 55. But the key point here, Your Honor, and how we know the RFC 2616 definition is what the specification is using, is it says we are going to designate this communication device to work as a client, designate this communication device to work as an agent. The labels that are provided in figure three is, what is the function being performed at that time? Which particular of the three software modules, all of which run in parallel, is operating at that time? And that's the label we're going to give it. And that's exactly what RFC 2616 says. It says, look, a device can and should provide multiple roles. We're going to look at the particular connection to see what the definition is, and what is the function as implemented in software that is being done. So RFC 2616 is entirely consistent with how the specification uses these labels. What did you think of the analogy? So I thought the analogy supported your side. I don't know if I got an answer per se from opposing counsel on the analogy, but what do you think of that? Your Honor, I agree with your analysis. My co-counsel, Mr. Breaux, also agrees with it. I'm having trouble understanding the analogy, period. I think the bottom line is, you know, analogies are always tough, but the bottom line is, this is a computer science application where everyone looks at the role. I think what's incredibly important here, Your Honor, is that there is no distinction in hardware in this specification. I do want to address that counsel stated that what makes this so great is that before this patent, there were never personal computers that acted as a proxy. Respectfully, that's not the finding of the PTAB. Crowds, which is one of the references in four of the, excuse me, six of the eight petitions, describes each of the user computers, user computers of the users, exact same thing as consumer computers, acting as a proxy, that last hop before the web server. Plamondon is even more explicit. Plamondon states that the Appliance 200, which is the last stop, can be a laptop, it can be a mobile device, it can be a personal computer. So the story of we invented the idea of having a proxy that is a personal computer, that does not, that would not ultimately flip any of these eight IPRs. Can I take you back to lexicography? Yes, Your Honor. In column two, we have said, I think at least in Kiyosera, that the phrase referred to herein as is lexicography, or at least it's a strong indication that it's lexicography. I take it you're agreeing that it is lexicography in this patent for purposes of describing figure two, at least? Yes, Your Honor, I think that's exactly correct, which is, I think the case law also says that you look at the entire surrounding language and what did they actually say, and they did not, here they did not say the words client device refer to a computer of consumers. They said in this thing. Though they don't ever say as referred to herein only means, when we're talking about prior art, or only means figure two, you would acknowledge they're not that clear in saying it doesn't carry through to their claims. No, Your Honor, I would disagree with that. I think they are that clear, because the very beginning of the sentence says, in the network 50 comma, files are stored on computers of consumers referred to herein as client devices 60. So it started with in the network 50. It did not define the word client devices. It defined a specific object, client device 60. And Your Honor, I do want to just point out that both the PTAB and the district court pointed out that assume there is lexicography, that they're still looking at it wrong. Because what is being described here is if you word search this patent for the word consumer, what is being talked about is not a user, not a person sitting at their personal laptop. Consumer is consumer of content, which is the opposite of a provider of content. And so even if there was lexicography, a consumer of content, it's not this hardware distinction they want to make. In trying to figure what a person of ordinary skill in the art would read into this sentence that's arguable, or I think agreed upon lexicography for at least a limited purpose, the fact that it appears in the background, might that lead one of ordinary skill in the art to think, well, this is going to carry all the way through to the claims because this is sort of foundational to what I, a person of ordinary skill in the art, would, of course, understand about a client device, all client devices in this art. That is, the fact that it's in the background, doesn't that cut against you? I would again say no, Your Honor, for the reason that the word client devices does not appear in the descriptions of any of the embodiments of the invention. So you have to skip forward many years in time until the word client device appears in these continuation claims. So in reading the specification as a whole, the inventors chose to use a different word, communication device, in conjunction with all the embodiments of the invention. And, Your Honor, I do want to point something out just very briefly. If the court were to say, look, there's a hardware distinction based on lexicography as to client device, again, that does not change the result of any of these IPRs because it is not the other shoe falling that appellant would then be entitled to a hardware distinction based on server. And what we see, Your Honor, is the only way applying their definition of server that appellant was able to even show that concept was to create a figure that is not actually in the patent. What they are asking for in terms of having a definition of server being not a client, not client hardware, not personal computer hardware, that is directly contrary to every disclosed embodiment in this patent, which describes that everything in Figure 3 is generic components of computers. And so they do not rise and fall together. There is a separate issue, not supported by lexicography, as to the server component, and that is where they actually have to win to overcome crowds, morph mix, and claim conduct. They have to win on both claim constructions. Yes, they do, Your Honor. And I just think importantly that the evidence to support the server construction is substantially weaker. It relates to the idea of, well, these have to be opposite terms, which the preferred embodiment, the only embodiment, is directly contrary to that. When in the gray brief, counsel admitted that if you look at Figure 3, the only figure, and you apply their constructions, it does not match Figure 3. It only matches a figure that their expert created. And when their expert created that figure, he did not say to a single portion of the specification. He just said, well, you could put this there. Just real quick, in columns four and five, maybe elsewhere, we talked about this with Judge Cunningham, the reference to the device may serve as a client, peer, or agent, sort of interchangeably. There's no reference to server, I don't think, in those statements. Is that significant? It is not, Your Honor, because if you look at the idea of what's a server function, what's a client function, the peer function is I get a request for content, and I provide you content, which means peer is server, agent does both, client and server, I apologize for speaking quickly, I don't want to run out of time, and the client is the role of a client. So those three things cover both the client and the server role. None of them talk about hardware at all. And what the board and the district court correctly found is it may not have listed it as an example, but a negative exclusion is a huge burden, and they don't come close for that negative exclusion of not a server. Thank you. Thank you. Thank you, Your Honor. Mr. Harkins, you have three minutes. Thank you, Your Honor. This issue of the roles devices play versus what are the devices is the core issue of what these patents are about. And there was some confusion in the district court, and I think very much so in the board's findings, about the difference of roles versus the difference of devices. There were four different claim construction orders that Magistrate Payne issued regarding the same claim language. Four of them. He kept revisiting it. But what happened at the end of that process was that he, and it was a case called NetNet, was the last ruling that he got to make on this, was that he said, look, I'm going to find that a client device is a communication device, and a server is not. A server is not a communication device. That you make a good point, and the point that we argued was, if the appellants were right, what is the difference between figure one, the prior art, and figure three? What is the difference? Figure one has a client going to a server, going to a web server. Figure three has a client going to an agent, going to a web server. They do the same thing. So looking at your statement of issues in your blue brief, I think you've identified four different issues for this Court. Right. If we rule against you on issues one and two, do we need to reach issues three or four? We do, Your Honor, and the reason is because if you are to affirm the board's constructions, they still wouldn't invalidate the claims. We'd still have a problem there, and there are still obvious misconsiderations for some of these that would require you to look at the art. Am I... I want to make sure I'm looking at the four issues right now. Sure. Yeah, let's get on the same page. Okay. Yes. I spoke correctly, which is the issue of invalidity and the issue of secondary considerations is still an issue, even if you adopt the board's construction, and the reason is because not only, and I've said that there's prosecution history and there's a lexicography, but also the claims themselves have a problem. The board said a client devised a claim that a device is a device that is operating in the role of a client and a server is a device that is operating in the role of a server, but the claims say that you have a client device that is sending files in response to requests. That's not operating in the role of a client. So what they pointed out in this prior art is not met. It's an impossibility. It's a thing you can't meet, which is if you've got a John Doe in the crowds reference and that is sending responses to requests, that is not operating in the role of a client under the same standards we're all talking about because that's not a client role. Okay, Mr. Harkins, you've run out of time. Okay. The case is submitted. Thank you.