[00:00:00] Speaker 04: We will hear argument next in number 241571, Akamai Technologies against MediaPoint. [00:00:08] Speaker 04: Mr. Liu. [00:00:10] Speaker 00: Good morning, Your Honors. [00:00:11] Speaker 00: May it please the Court? [00:00:13] Speaker 00: Larry Liu, on behalf of MediaPoint, Inc. [00:00:15] Speaker 00: and AMHC, Inc. [00:00:17] Speaker 00: I'd like to reserve three minutes for rebuttal. [00:00:22] Speaker 00: The District Court erred in three ways. [00:00:24] Speaker 00: First, at claim construction. [00:00:27] Speaker 00: by issuing a case-dispositive discovery sanction, and by granting summary judgment by construing factual instances against media point and non-movement. [00:00:38] Speaker 00: This court should reverse the media point. [00:00:40] Speaker 00: Let me start with claim construction. [00:00:43] Speaker 00: The patents I've issued, the 426 and 195 patents, concern the delivery of content across the internet using the best and often most droughts. [00:00:55] Speaker 00: The district court erroneously held the best and optimal terms indefinite. [00:01:01] Speaker 00: That was error because the claims expressly describe how to determine the best and optimal routes. [00:01:07] Speaker 00: That is, by mapping and comparing trace routes. [00:01:12] Speaker 00: claim one of the 426 patents. [00:01:14] Speaker 04: So would the following be a fair description? [00:01:17] Speaker 04: That the claims certainly say you do something with the routes, but they don't say exactly what you do with them. [00:01:27] Speaker 04: That there can be a use in the decision-making process, but the claims don't specify [00:01:35] Speaker 04: the determinative factors for making the decision itself, whether it's number of hops or latency or reliability or possibly other considerations as well. [00:01:49] Speaker 04: And it's that problem that undergirds the district court's indefiniteness determination. [00:01:57] Speaker 00: Yes, Your Honor, but the prioritization problem is not a problem at all for indefiniteness. [00:02:01] Speaker 00: That's really a problem of implementation. [00:02:04] Speaker 00: Indefiniteness is about [00:02:05] Speaker 00: the reasonable certainty of the claim scope. [00:02:08] Speaker 01: Isn't it really about whether somebody could look at the claim and understand what falls within it so they can know whether they're going to infringe the claim or not infringe the claim? [00:02:22] Speaker 01: And when something says best or optimal and the specification discloses lots of different possibilities for determining that, it makes it hard to understand what [00:02:34] Speaker 01: What falls within the claim and what does not? [00:02:38] Speaker 00: You're right Judge Sullivan, that's the purpose of indefiniteness. [00:02:42] Speaker 00: There's no issue here because the claim language makes clear that the best and optimal routes are determined by mapping and comparing trace routes. [00:02:50] Speaker 00: Here, whether one prioritizes hops or latency, either one would infringe, so long as the reported infringer determined the best or optimal routes by mapping and comparing the outputs of trace routes. [00:03:04] Speaker 00: To your question, Judge Feranto, as to whether there are other factors, like reliability or potentially other factors, those are not an issue because the district court expressed and construed trace routes to output only number of hops and latency. [00:03:18] Speaker 00: And that claim construction is not challenged on appeal by either party. [00:03:23] Speaker 02: Off my- It might be instructive for us to actually look at some of the claim language, because you're contending that [00:03:29] Speaker 02: It is clarifying in this claim language and you would know what falls within the scope of the claim or what falls outside of it. [00:03:36] Speaker 02: Can you walk us through some of the pertinent claim language that you think gives that level of certainty? [00:03:41] Speaker 00: Yes, Judge Cunningham. [00:03:42] Speaker 00: So claim one of the 426 patent discloses a, quote, mapping engine that is configured to map trace routes so as to determine one or more optimal routes. [00:03:56] Speaker 00: So claim one is expressly drawing a causal connection [00:03:59] Speaker 00: between the mapping of the trace routes and the optimal route. [00:04:04] Speaker 00: Claim 17, which Occamite admits is a, quote, narrower claim, likewise says, determining a best route, quote, based on a comparison between the trace routes. [00:04:17] Speaker 00: It says nothing else about any other factor, and Occamite tries to muddy the water by citing [00:04:23] Speaker 00: parts of the specification that are not tethered to the best or optimal terms, either as to best or optimal route or best and optimal network link. [00:04:36] Speaker 04: You'll correct me so my recollection is as following is as follows that In your gray brief you make a point like this But maybe not so much in your blue brief that in your blue brief you make the argument that it's clear from the patent that there's [00:04:55] Speaker 04: everybody would kind of know hops and latency or the main thing or something and they pretty much go together though not always and that's good enough. [00:05:04] Speaker 04: Not that I don't remember from an argument that [00:05:10] Speaker 04: As long as trace roots are used, you know if you're a potential infringer that you're infringing, regardless of what you then do with them except use them. [00:05:23] Speaker 00: You're right, Judge Franta. [00:05:24] Speaker 00: That argument is not crystal clear in our blue brief. [00:05:27] Speaker 00: In our blue brief, it's raised expressly in our gray brief at five in responding to optimized arguments about prioritization. [00:05:34] Speaker 00: We argue in the blue brief that there is no record evidence that the two factors, that is hops, [00:05:40] Speaker 00: and latency would ever diverge. [00:05:43] Speaker 00: And that is Optimize's burden to prove by clear and convincing evidence under this Court's precedent in Guadalupe Allison to show that if there are multiple methods for determining what is optimal or best, here Optimize's contention is that's a prioritization between hops and latency. [00:05:59] Speaker 00: That's their burden to show that those factors would ever diverge. [00:06:04] Speaker 00: There is no revenue. [00:06:05] Speaker 02: But by raising it in your gray brief, are you raising it too late, that specific argument? [00:06:09] Speaker 00: No, Your Honor, because it's another argument in support of the same issue as to whether this claim term is definite or indefinite. [00:06:18] Speaker 00: Occamite has an opportunity to respond here before this court as well here at argument, so there's no prejudice to it as to our intention that there's no waiver. [00:06:28] Speaker 00: So whether or not one prioritizes hops or latency [00:06:34] Speaker 00: the claim term is infringed, and that solves the indemnity. [00:06:38] Speaker 04: Can you remind me, since this came up under claim construction, did you, I think I remember you did, but what was your proposed claim construction for this language? [00:06:49] Speaker 00: Our construction? [00:06:50] Speaker 04: Where would I find it? [00:06:52] Speaker 00: It's in the appendix. [00:07:02] Speaker 04: Okay, okay, I can look there Appendix page 10. [00:07:10] Speaker 00: Yes So as to the best and optimal routes our construction there was to distinguish those two terms from each other and simply make clear that one Allows for the potential of one or more Most efficient or optimal routes and best refers to the singular best routes [00:07:30] Speaker 00: And what occurred in the briefing below was that MediaPoint responded to optimize indefiniteness arguments and explained, as it does on appeal, that the mapping and comparison of trace routes resolves the supposed subjectivity problem. [00:07:45] Speaker 01: Your interpretation here at page 810 seems quite a bit different than what you're arguing now. [00:07:51] Speaker 00: No, Your Honor, because the issue when we were construing those claims [00:07:56] Speaker 00: was simply to distinguish best and optimal. [00:07:58] Speaker 00: In the briefing, as we argued, we argued that those terms are not indefinite and are subject to the mapping of traceroutes. [00:08:07] Speaker 01: That's precisely- Now you seem to be focusing on the interpretation here is path determined to be among the most favorable based on one or more characteristics, and now you're identifying what those characteristics are. [00:08:21] Speaker 00: Yes, Your Honor. [00:08:22] Speaker 01: So how is that not different? [00:08:23] Speaker 00: Those characteristics are [00:08:25] Speaker 00: determined by the mapping of trace routes. [00:08:27] Speaker 00: One or more characteristics, and then the relevant claim language that follows optimal in the claims, claim one of the 426 patent is, sorry, in the claim one it proceeds a mapping engine that is configured to map trace routes. [00:08:44] Speaker 00: so as to determine one or more optimal routes. [00:08:46] Speaker 04: Maybe a different way to get at this might be this. [00:08:50] Speaker 04: Why isn't what you're suggesting one that effectively reads optimal out or at least doesn't provide, doesn't treat it as independent? [00:09:02] Speaker 04: You're suggesting that a mapping engine that is configured to map trace routes so as to determine a route [00:09:10] Speaker 04: Or optimal route would be the same as so as to determine a route and the word optimal is what was the subject of this this claim construction so that's got to add something and the at the addition of something is I think what is Was held to be indefinite More than just if you use these then you are in the claim. [00:09:37] Speaker 04: That's not [00:09:38] Speaker 04: You have to use these to then do something with them that amounts to following this other standard of optimal, and that is the standard that is left too uncertain. [00:09:53] Speaker 00: Well, it's not, Your Honor, because it simply allows, it still thinks that the term optimal is constrained by trace tracks. [00:10:00] Speaker 00: It's not superfluous because it allows the procedure to know [00:10:05] Speaker 00: You can either prioritize hops or you can prioritize plates. [00:10:08] Speaker 04: But to say it's constrained may be a way of saying there are two requirements that have to be met to come within it. [00:10:14] Speaker 04: One, you have to use these things. [00:10:16] Speaker 04: Second, it has to be optimal. [00:10:18] Speaker 04: And we'll stipulate that the first is definite. [00:10:22] Speaker 04: You're using these things. [00:10:23] Speaker 04: But the second is, in the conjunctive, a requirement, not a conjunctive in here, in my translation of this into a conjunctive with an ad, is also a requirement, and that's what's too uncertain. [00:10:36] Speaker 00: Right, Your Honor, but I would push back on the suggestion that those two things are conjunctive in the claim language. [00:10:43] Speaker 00: The claim language instead draws a causal connection between the two. [00:10:46] Speaker 00: It doesn't say map trace routes and determine an optimal route. [00:10:51] Speaker 00: It says mapping trace routes so as to determine one or more optimal routes, and determining a best route based on a comparison between trace routes. [00:11:02] Speaker 04: But then aren't you giving no meaning to the term optimal if whatever the route is that's covered is just any route that you have, not route, the choice. [00:11:17] Speaker 04: Yeah, the route that you choose follows from using the traceroutes. [00:11:25] Speaker 00: No, Your Honor. [00:11:26] Speaker 00: It still gives a fact of the claim term, because it fetches the purpose of the patent, which is to determine the most efficient network path. [00:11:33] Speaker 00: It makes clear that the patent they've proceeded can still choose between any of those factors that are output by traceroutes. [00:11:41] Speaker 02: What is your best [00:11:42] Speaker 02: case support for the language here being sufficiently definite, so that a person would actually know what falls within the scope of the claims at issue here. [00:11:56] Speaker 00: So Your Honor, I think this case actually follows all four shiori from this court's precedent, Sonic Tech and Guangdong Allison. [00:12:02] Speaker 00: This court has never considered a term of degree where the claim language itself defines the method by which the term of degree is bounded. [00:12:12] Speaker 00: In Sonic's tech, the issue was visually negligible, and the court considered examples in the specification that set the outer bounds and provided a, quote, objective baseline for what was visually negligible. [00:12:27] Speaker 00: Here, the claim language itself does that, and the examples in the specification give effect to that. [00:12:34] Speaker 00: That's the comparisons in tables D, E, F, and G of the specification. [00:12:40] Speaker 00: That's Appendix 61 and 62. [00:12:42] Speaker 00: in comparing the routes between note one and the client and note two and the client. [00:12:48] Speaker 02: Do you agree that there's no meaningful difference between the representative claims identified by the parties that should affect our analysis in this case? [00:12:57] Speaker 00: There is no meaningful difference between the claim one and claim 17 in the 426 pattern that affects the claim construction. [00:13:03] Speaker 00: I agree. [00:13:05] Speaker 02: What about briefly talking about the non-infringement determination here? [00:13:10] Speaker 02: Does exclusion of the DNS resolver and consistent hashing theories affect the non-infringement determination? [00:13:18] Speaker 00: No, Your Honor. [00:13:19] Speaker 00: As we argued below to the district court, neither of those allegedly new theories we argued of those were in fact bootstrapped to the exclusion of Mapper. [00:13:28] Speaker 00: But neither of those were claimed or case dispositive. [00:13:31] Speaker 00: The exclusion of the Mapper allegation in the extradural court was treated as case dispositive. [00:13:40] Speaker 04: I think my main takeaway from your argument about the striking, or at least I will say that right now, [00:13:54] Speaker 04: sense of the strongest point you make is the other side has never identified any concrete prejudice. [00:14:00] Speaker 04: So I want to put that aside. [00:14:01] Speaker 04: We don't have to get to that if we think the summary judgment on the merits is correct, right? [00:14:07] Speaker 04: So what is your best argument for why the district court was wrong about the summary judgment of non-infringement on the merits? [00:14:17] Speaker 00: Sure, Your Honor. [00:14:17] Speaker 00: So let me table set on that issue and start with a factual record. [00:14:23] Speaker 00: It's undisputed that there are two documents, one document that shows the diagram at 69 of the blue brief, Akamai's mapper receiving a request for content. [00:14:32] Speaker 00: Another Akamai internal document describes that mapper's job is to match the request for content with the best edge server to handle that request. [00:14:42] Speaker 00: And so for the district court to rule otherwise, it could only have done so on two grounds. [00:14:48] Speaker 00: Either those documents were as a matter of law [00:14:53] Speaker 00: insufficient to raise a genuine issue of material fact. [00:14:56] Speaker 00: Or the district court was implicitly construing that term, request for content, as limited to a particular kind of computing message. [00:15:06] Speaker 00: And either one of those would have been error. [00:15:09] Speaker 00: On the facts, for the district court to reach that conclusion, it would have had to draw inferences against the non-movement media point. [00:15:16] Speaker 00: And you see that clearly in the district court's decision, which is it simply calls those documents, quote, too high a level. [00:15:23] Speaker 00: the concessions from MediaPoint's expert were only concessions under optimized construction of requests for content. [00:15:33] Speaker 00: And so to give an analogy, MediaPoint's expert explained that MAPR [00:15:38] Speaker 00: does not merely receive a DNS query divorced from context, much like a doctor's office receives a request for medical services when it receives an appointment request, even if it doesn't know the specific procedure. [00:15:54] Speaker 04: So this may be wrong, but let me just sort of tell you what is currently in my mind about the picture. [00:16:02] Speaker 04: I can imagine a system in which I'm a user, I want some content, I send a message up to device number one, that message specifies the content, the actual content, but device number one then sends a message up to device number two and it doesn't say what the content is, it says translate the beginning of the URL [00:16:25] Speaker 04: into a IP address without saying what the content is. [00:16:30] Speaker 04: So device number two never gets any message of any sort that has information about the specific content because device number one has stripped out all of that and just said I want to know where [00:16:43] Speaker 04: an IP address for www.netflix.com. [00:16:48] Speaker 04: No more. [00:16:50] Speaker 04: Is it right, or Mr. Saunders maybe will tell us, I guess I sort of understood that what the district court was saying was what I just described [00:17:00] Speaker 04: is the only evidence, is what the only evidence shows their system to be. [00:17:06] Speaker 04: Device number two is mapper. [00:17:08] Speaker 04: It never hears about the content. [00:17:10] Speaker 04: Something else strips that out. [00:17:13] Speaker 04: Address that picture, whether it's incorrect, [00:17:17] Speaker 00: That's roughly accurate, Judge Toronto, that Mapper never knows what the specific content is in the sense it doesn't know whether you requested the particular, let's say, Spider-Man movie or the basketball game tonight. [00:17:31] Speaker 00: Our point is the request for content, though, is the user's request that's entered at his or her device or browser. [00:17:39] Speaker 00: But but I thought the claim says that the relevant I forget what the claim language is the relevant server has to receive that it says The management center has to receive an initial request for media content right our point is it receives that request by Receiving the DNS query and in doing so it understands that the user has requested content [00:18:04] Speaker 00: And that's supported by the fact that the claim language, the rest of the claim elements, use the term content stream, as opposed to the content itself, or content request. [00:18:15] Speaker 00: So our point is the request for media content is a plain and ordinary, meaning construction of that term. [00:18:22] Speaker 00: It's a colloquialism. [00:18:23] Speaker 00: It's the user's request entered at his or her browser, not a specific computing message. [00:18:29] Speaker 00: And so in framing Occamized, [00:18:32] Speaker 00: argument, Judge Stronto, would be to accept if your framing accepts their framing of request for media content as a particular computing message. [00:18:40] Speaker 00: Our point is the mapper doesn't have to receive [00:18:42] Speaker 00: an HTTP request or any type of particular request that identifies the specific content for there to be infringement of that element. [00:18:50] Speaker 04: That it's enough that, in my example, device number two gets a request for translation into an IP address of a domain name, and the only time it ever gets that request is if somebody has requested content, and so it is effectively receiving a request for content, though only a portion of it. [00:19:12] Speaker 00: That's precisely right, Your Honor, except for I would excise to accept a portion of it. [00:19:17] Speaker 00: I would agree with every other part of your characterization. [00:19:21] Speaker 00: I see him way over my time, Your Honors. [00:19:23] Speaker 04: It wasn't your fault. [00:19:24] Speaker 04: Why don't you sit down and we'll hear from Mr. Saunders to get your rebuttal signed back. [00:19:34] Speaker 03: May it please the court, starting with indefiniteness, there are two problems with media points condition. [00:19:40] Speaker 03: One is we have, and we know based on the specification, that we have this open-ended long list of factors. [00:19:46] Speaker 03: And they really don't have any argument to consider that whole list of factors. [00:19:52] Speaker 03: But second, even limiting it solely to the output of traceroutes, there still is this objective determination that needs to be made here. [00:20:01] Speaker 03: Traceroutes will tell you the latency, [00:20:03] Speaker 03: the hops, the reliability. [00:20:06] Speaker 03: But what is optimal? [00:20:07] Speaker 03: What is best? [00:20:08] Speaker 03: It depends on what you prioritize. [00:20:10] Speaker 03: It's the car example we have in there. [00:20:12] Speaker 03: What is best between a sports car, a pickup truck, a hybrid vehicle? [00:20:17] Speaker 03: It depends on, do you want speed? [00:20:18] Speaker 03: Do you want hauling capacity? [00:20:21] Speaker 03: Do you want energy efficiency? [00:20:23] Speaker 03: And so, in this context, we started, as Judge Stoll pointed out, with very broad, [00:20:32] Speaker 03: claim constructions offered by MediaPoint. [00:20:35] Speaker 03: They're really circular and subjective. [00:20:38] Speaker 03: Even with this sort of new attempt to limit it to trace routes, the way it's being described here, as Judge Toronto's questions were probing, you could just cross the word optimal, cross the word best out of the claims, it would mean the same thing. [00:20:53] Speaker 03: It has to mean something, and the problem is it's supposed to be an interpretation, but [00:20:59] Speaker 03: that you don't have the guidance here to know how to make that balance, even focused on trace routes. [00:21:05] Speaker 01: One thing that you said that's a little bit different is you said it could be hops latency plus reliability. [00:21:12] Speaker 01: What is the difference? [00:21:13] Speaker 01: Why do you bring in reliability under trace routes? [00:21:16] Speaker 01: as opposed to what I'm hearing from the other side. [00:21:21] Speaker 03: So to be clear, they conceded in the district court that reliability would be part of it. [00:21:27] Speaker 03: So appendix 2030, their Markman reply, as the specification explains, traceroute calculate route length, transmission reliability, and latency. [00:21:38] Speaker 03: But let me give you sort of a concrete example. [00:21:40] Speaker 03: We've been talking in the brief about table A, which is one of the circumstances where earlier on in the hops, you actually have a higher round-trip latency from the beginning to that earlier hop than you do from the beginning to the end. [00:21:58] Speaker 03: And that can be a signal of a reliability problem, which is as you were doing these trace routes, at some point you were actually getting a longer time. [00:22:08] Speaker 03: So your final result at the destination, maybe that looks like a good time, but the fact that earlier on something was going on there, there was a jitter, it was temporarily slower, may make you think, is there a reliability problem? [00:22:26] Speaker 03: in this route. [00:22:27] Speaker 03: So even if I like that bottom line, how it's getting to the client, what's going to happen when I have to run this again? [00:22:34] Speaker 04: Just to make sure I'm in the same place, this is because 189 comes after but is smaller than 217? [00:22:41] Speaker 03: Correct. [00:22:41] Speaker 03: And so we're all clear, these are round trip times. [00:22:46] Speaker 03: So each time on the trace route, it's from beginning to that point. [00:22:49] Speaker 03: It's the beginning of the point. [00:22:50] Speaker 03: They're not cumulative. [00:22:52] Speaker 03: So it is showing, you know, [00:22:56] Speaker 03: Oh, 189. [00:22:57] Speaker 04: So the fifth test path, which goes through all the others, is actually from beginning to fifth one, and that one was faster. [00:23:07] Speaker 04: When it did the previous test path only to number four, it was longer, but by the time the fifth test was run, there was smooth sailing. [00:23:19] Speaker 03: Absolutely. [00:23:22] Speaker 03: to balance, because that's a reliability question. [00:23:24] Speaker 03: Do I care about this? [00:23:26] Speaker 03: Am I content to say, oh, I'm going to optimize based on latency at the end? [00:23:32] Speaker 03: Or do I say, actually, I'm worried about those higher numbers earlier, and when I do this one more time to try to reach the end, am I going to have that same jitter? [00:23:41] Speaker 03: Is it going to be a problem? [00:23:42] Speaker 03: So even within traceroutes, these balancing of factors doesn't leave the objective boundaries [00:23:50] Speaker 04: So why is this materially different from Sonics? [00:23:57] Speaker 03: So in Sonics, the claim was talking about what is visually negligible. [00:24:06] Speaker 03: And there was really only one dimension there. [00:24:09] Speaker 03: And the court said, that's an objective standard because it's based on the capacity of the human eye, which is something that can be objectively determined. [00:24:20] Speaker 03: And so based on that, you could say, OK, we know where the cutoff is. [00:24:28] Speaker 03: Really, where the average person can no longer see this, then that becomes visually negligible. [00:24:34] Speaker 03: That's fundamentally different from an optimal claim like the intellectual ventures case, where what has to be optimized based on different factors ends up being what factors do you prioritize? [00:24:48] Speaker 03: For the other issues, because it was just first of jumping straight to summary judgment, they have to win both on exclusion and summary judgment. [00:25:00] Speaker 03: The key thing to remember here is what the district court is focusing on first and foremost is their expert's own testimony. [00:25:07] Speaker 03: And what is actually going over to MAPR? [00:25:13] Speaker 03: It's the DNS query. [00:25:15] Speaker 02: Just in a rough one thing, you said they have to win on both exclusion and summary judgment. [00:25:19] Speaker 02: Do you think they could win on summary judgment on the merits? [00:25:23] Speaker 02: And I think others with questions, Judge Toronto asked, even if we didn't agree with respect to the exclusion that was made. [00:25:30] Speaker 03: I'm saying we win on either. [00:25:36] Speaker 03: We went on summary judgment. [00:25:38] Speaker 03: Even reaching the merits, we went on. [00:25:40] Speaker 03: So you never need to consider the exclusion. [00:25:44] Speaker 03: And on this question, so we're agreed, it's just the DNS query going over. [00:25:50] Speaker 03: So what did their experts say at 2770 through 2771? [00:25:55] Speaker 03: Quote, a DNS query itself is not the request for content described in the 195 pact. [00:26:03] Speaker 03: And the reason is, as Judge Toronto's questioning was getting at, [00:26:07] Speaker 03: that whatever the user is doing, we have a claim that requires that the request for content be received. [00:26:17] Speaker 03: And what the DNS query does, it has information stripped out. [00:26:21] Speaker 03: It doesn't know, if I type in sort of a web address and I'm looking for a specific video, whatever I'm doing there, that information and the actual content I'm looking for never makes it to Akamai's map. [00:26:36] Speaker 03: All that ends up going is the host name. [00:26:38] Speaker 03: It just wants that translated into the numerical IP address. [00:26:42] Speaker 03: That comes back. [00:26:44] Speaker 03: And then I know what edge server to go to, and the request for content is made there. [00:26:51] Speaker 03: So Akamai system doesn't know at the level of the mapper what content it is. [00:26:55] Speaker 04: And on the evidence here, do I understand correctly, we don't have [00:27:03] Speaker 04: proof of a system in which the central management server, the mapper, looks at what content is being requested to determine which of a number of servers to refer to send back as the place to go. [00:27:23] Speaker 03: Yeah, absolutely not. [00:27:24] Speaker 03: I mean, the Akamai system works based on serial numbers, so think of who is the [00:27:30] Speaker 03: the client for them would be the third party whose content they're hosting. [00:27:35] Speaker 03: And you would have a whole bunch of different companies or sites may get lumped together under the same serial number. [00:27:42] Speaker 03: Think of it as all the local businesses in this region might be under the same serial number. [00:27:49] Speaker 03: And so a request will be sent to an edge server that might be local to this region, but it doesn't have anything to do in that system, in that first [00:28:00] Speaker 03: sort of routing with what content's being requested, because the mapper doesn't know what content you're requesting. [00:28:08] Speaker 03: You know, in terms of the diagram, so there is a figure that was being sort of misinterpreted to say, oh, is the request for content. [00:28:18] Speaker 04: Is this the one where there are two different versions? [00:28:20] Speaker 03: Yes. [00:28:21] Speaker 03: Yeah. [00:28:21] Speaker 03: And so the key context is at 2659 of the appendix, you can see the questions that were asked [00:28:37] Speaker 03: And our expert at 2688 says it doesn't operate in the way that he was interpreting that figure. [00:28:47] Speaker 03: The figure then gets revised. [00:28:48] Speaker 03: This is 2683 of the appendix. [00:28:51] Speaker 03: And their expert admits that they were wrong. [00:28:54] Speaker 04: I'm sorry. [00:28:54] Speaker 04: Passive voice. [00:28:55] Speaker 04: Figure then gets revised. [00:29:01] Speaker 03: It said, if you think this figure is showing that, clearly this figure isn't clear enough. [00:29:09] Speaker 03: And it revised it. [00:29:10] Speaker 03: And then importantly, their expert, this is at 2683, agrees that the revised figure is, quote, consistent with how a DNS query would work in the real world. [00:29:23] Speaker 04: Right, so they can't have a and that's and that's quite crucial because otherwise it might be a fact question Which of the two versions you produce you should be stuck with or benefit from? [00:29:34] Speaker 03: Yeah, so but you can't get to a fact question By saying we want to look at the old figure rather than how it actually operates and unless there's a dispute about whether the first figure [00:29:51] Speaker 03: And here, on top of the sort of testimony about the figures, direct testimony from their expert as to what is going over. [00:30:00] Speaker 03: And we heard it today as well. [00:30:01] Speaker 03: It's only that DNS query. [00:30:04] Speaker 03: And the direct admission that the DNS query is not the request for content. [00:30:10] Speaker 03: I'm happy to address the exclusion point, but this court doesn't. [00:30:14] Speaker 04: I guess I'd be interested. [00:30:15] Speaker 04: I mean, you heard me say before that I think there's a hole in what I've read from you on the subject of prejudice. [00:30:27] Speaker 03: Right. [00:30:27] Speaker 03: So let me be more direct. [00:30:29] Speaker 03: The reason for that is because we think that [00:30:34] Speaker 03: going through eight months of discovery on the wrong theory and then be confronted with a new theory a month before your expert report is due is itself prejudice and then nothing more. [00:30:48] Speaker 04: Well one of the, I may be misremembering this, so I don't remember your [00:30:56] Speaker 04: ever saying or maybe more importantly the district court ever saying that he in your case should in his case would strike have stricken this if this came in the June final infringement contentions as opposed to nearly two months later in the August expert report. [00:31:19] Speaker 04: So I'm not sure [00:31:20] Speaker 04: eight months is the right time period to be thinking about. [00:31:26] Speaker 04: I would have thought that the two-month difference is so that... I think you understand my point. [00:31:33] Speaker 03: I understand your point. [00:31:34] Speaker 03: It's a very critical two months, because what happens in those two months is discovered and closed. [00:31:38] Speaker 04: Fact closes, claim construction comes out both, right? [00:31:42] Speaker 04: Doesn't claim construction come out in the middle in July? [00:31:44] Speaker 03: Before the contentions, if I'm remembering correctly. [00:31:50] Speaker 03: Okay, so that's okay And so you know fact discovery then closes and then we have an expert report with this this new theory New source code as we point out, you know an element 1b all the source code that's being identified in the final infringement contentions [00:32:12] Speaker 03: None of those files are mentioned anymore. [00:32:15] Speaker 03: We have new files that are being accused. [00:32:18] Speaker 03: But just to speak to your specific concern, we don't think when the burden's on them to prove harmlessness, [00:32:27] Speaker 03: that we have to come forward and say, oh, here is the specific deposition question we would have asked. [00:32:34] Speaker 03: But let me give an example in the context of this case that may help assuage that concern, which is one of their new theories was the claims have trace routes to the client. [00:32:47] Speaker 03: And they came in with the new theory about the DNS resolver and said, well, our contentions were doing trace routes to the client. [00:32:55] Speaker 03: Now we're going to do traceroutes to your third party DNS resolver, and then we're going to use that as a proxy. [00:33:02] Speaker 03: We're going to approximate from there. [00:33:05] Speaker 03: Knowing that claim, we may have done different prior art searches. [00:33:10] Speaker 03: If that is their view of the claim, that unlocks potentially different system art. [00:33:16] Speaker 03: Now, we haven't run those searches that the case was over, it was struck. [00:33:21] Speaker 03: But that's the type of thing where... Go ahead. [00:33:24] Speaker 02: Do you agree that exclusion of the DNS resolver and consistent hashing theories [00:33:30] Speaker 02: or do you believe the exclusion of the DNS resolver and consistent hashing theories affects the non-infringement determination here? [00:33:37] Speaker 03: Yes, it does, because they have to prove infringement on every element. [00:33:42] Speaker 03: And so if their theory for that traceroute to the client is now a new theory that's about approximating it based on the DNS resolver and that's excluded, then that would be disruptive for them as well. [00:33:57] Speaker 03: But in the mapper theory, [00:34:00] Speaker 03: as well was a new theory. [00:34:02] Speaker 03: We see that most dramatically in this shift in the source code that's being cited. [00:34:08] Speaker 02: Do you also agree if the district court did not actually make a finding regarding willfulness, fault, or bad faith? [00:34:14] Speaker 03: It did not make a direct express finding. [00:34:18] Speaker 03: We think implicit is a finding of fault. [00:34:20] Speaker 03: It was never asked to, and so that [00:34:23] Speaker 03: would be a forfeited argument here. [00:34:27] Speaker 03: And I do just want to clarify, in the Munchkin case, which is the Ninth Circuit case that we've cited, the reply brief came back and said, oh, there wasn't even an argument of harmlessness. [00:34:43] Speaker 03: It's looking at a sort of much [00:34:45] Speaker 03: earlier period, but for it's ECF 516 on that district court docket is the operative request. [00:34:54] Speaker 03: And it was talking about the substantially justified or harmless rule. [00:34:59] Speaker 03: Harmlessness is where there's no prejudice. [00:35:01] Speaker 03: So the point is munchkin is finding forfeiture of This specific finding that has to be made in a circumstance where a party is talking about harmlessness and substantial justification But as here doesn't say and you need to make this additional [00:35:18] Speaker 02: I know the parties made or identified different representative claims. [00:35:22] Speaker 02: Do you agree, though, that there's no meaningful difference between the representative claims identified the parties that should affect our analysis here? [00:35:29] Speaker 03: No, it shouldn't. [00:35:30] Speaker 03: The one thing I would say is on this issue, on indefiniteness, when you're looking at the whole range of factors that's given in the specification, [00:35:40] Speaker 03: they've attempted to distinguish that by saying, oh, that relates to the best performing node index. [00:35:47] Speaker 03: The node and the best node optimal traceroute, they're related, like the nodes are in the traceroute. [00:35:54] Speaker 03: And so I think the fact that [00:35:57] Speaker 03: they are treating a node claim as representative there is telling. [00:36:02] Speaker 03: And we have expert testimony in the record that says when you're optimizing for one, it's true for the other, and vice versa. [00:36:09] Speaker 03: I see my time as well. [00:36:11] Speaker 04: Thank you. [00:36:13] Speaker 04: You'll have your three. [00:36:20] Speaker 00: Thank you, Your Honors. [00:36:21] Speaker 00: I have two quick points on each of the issues here. [00:36:24] Speaker 00: Let me start with exclusion and work backwards. [00:36:28] Speaker 00: On prejudice, I heard my friend on the other side again lead with this idea that there's prejudice inherent in delay, but again, specifically not identify any prejudice. [00:36:39] Speaker 00: That is itself evidence that [00:36:41] Speaker 00: the theory did not fundamentally change. [00:36:44] Speaker 00: The theory was the same theory, and that's why they were able to proceed and not have to take any additional discovery, not identify any additional witnesses, and why they're unable to identify any prejudice here today. [00:36:58] Speaker 02: Well, I thought I just heard opposing counsel indicate that the prejudice is that by getting the theory at the later time, [00:37:07] Speaker 02: It meant that they didn't get to ask potential other questions they might have asked during the fact discovery phase, so to speak, because they weren't aware that the new theory was even going to be in play. [00:37:19] Speaker 02: So how do you respond to that? [00:37:20] Speaker 00: Well, Your Honor, it wasn't a new theory. [00:37:22] Speaker 00: And the only specific example that my friend on the other side hypothesized was about the DNS resolver to the client and not the core theory on which the district court granted the exclusion motion mapper. [00:37:34] Speaker 00: And MAPR was an identical functionality to GTN. [00:37:38] Speaker 00: We just simply made a semantic error that we diligently corrected. [00:37:42] Speaker 00: And so even the sort of question that he hypothesized he may have asked differently doesn't relate to the core of the district court's exclusion motion. [00:37:52] Speaker 00: On willfulness, fault, and bad faith, [00:37:54] Speaker 00: There's no question there was no finding here. [00:37:57] Speaker 00: My friend concedes it. [00:37:58] Speaker 00: And the district court did not make any finding that there was any disobedient conduct, which is a non-circuit standard under Liberty Insurance, R&R sales. [00:38:08] Speaker 00: It's not enough merely that the expert report isn't within the park's control. [00:38:12] Speaker 00: That would be true of any supposedly late disclosure. [00:38:15] Speaker 04: Does this point on your part depend on our accepting the proposition that this is non-forfeitable? [00:38:22] Speaker 00: No, Your Honor. [00:38:24] Speaker 00: It doesn't depend on that conclusion, it's not forfeitable, because we didn't forfeit it anyways. [00:38:29] Speaker 00: Because the willfulness, bad faith, and fault inquiry is subsumed under the substantial justification and harmlessness inquiry under Rule 37, and that was indisputably raised below. [00:38:40] Speaker 00: And if in any event, if there was forfeiture, this is a canonical case in which it should be excused because it's a pure question of the law. [00:38:48] Speaker 00: It's not even disputed in this case that there was no finding or that that finding is required. [00:38:53] Speaker 00: And so the reversal is straightforwardly required. [00:38:57] Speaker 00: On summary judgment, again, everything my friend has said depends on a claim construction of requests for media content as identifying a specific [00:39:10] Speaker 00: piece of content. [00:39:13] Speaker 04: Well, I guess the really simple version that I think I heard is your expert said that you agree that only the DNS inquiry, not anything specifically about the content, goes up to Mapper, and that your expert said that is not the request for content that the claim talks about. [00:39:37] Speaker 00: Two questions, Your Honor. [00:39:39] Speaker 00: So one, our expert is [00:39:40] Speaker 00: The point he was making is deposition, is that those two things are not definitionally the same thing. [00:39:46] Speaker 00: Because definitionally, the DNS query asks for an IP address, and a request for content is a user requesting content at his or her device. [00:39:54] Speaker 00: The deposition testimony that my friend omits, this is at the grade brief at 32. [00:40:00] Speaker 00: Dr. Rubin said. [00:40:00] Speaker 04: I'm sorry, do you have a JA site, the actual deposition? [00:40:03] Speaker 00: Yes, it's appendix 3691. [00:40:06] Speaker 04: Thank you. [00:40:09] Speaker 00: Dr. Rubin said, the user initiates the request for content, and that results in the DNS query being received by the mapper that is related to that user's request. [00:40:18] Speaker 00: And that is the mapper receiving the request for content when it gets that DNS query that relates to the user's request for content. [00:40:27] Speaker 00: And so that's consistent with the summary that you gave at the end of my initial time, Dr. Rubin, the judge for content. [00:40:36] Speaker 00: Finally, on claim construction. [00:40:40] Speaker 00: I'd like to make one final point in closing, which is, again, the prioritization issue is a complete red herring. [00:40:47] Speaker 00: Reliability, whether it's included or not, would be an inferable factor from traceroutes anyways. [00:40:55] Speaker 00: Whether it's two factors or three factors, it doesn't matter because the scope of the claim is clear. [00:41:00] Speaker 00: Whether you prioritize reliability, hops, latency, as long as you are doing so based on mapping of traceroutes, that would infringe the claim. [00:41:10] Speaker 00: And so this doesn't make it subjective, like intellectual ventures, where the specification expressly said quality of service depends on whatever the subjective vagaries of the user are. [00:41:21] Speaker 00: There's no such language in this specification tied to the best or optimal terms. [00:41:26] Speaker 00: And even if there were, it couldn't override the expressed claim language, which draws a causal connection between the best and optimal terms and the mapping and comparison of traceroutes. [00:41:37] Speaker 00: Thank you. [00:41:38] Speaker 04: Thank you. [00:41:39] Speaker 04: Thank you for both counsel and the cases submitted.