Oral argument — Arista Networks, Inc. v. ITC

versus the International Trade Commission, 16-2563. Mr. Quinn, how are you? I'm going to need some help here, how you've divided up your time, okay, just to make sure I get everybody in the right order. So, Mr. Quinn, you have 15 minutes, you're saving three for rebuttal, is that correct or not? I'm Mr. Powers, I'm saving six. Okay, Mr. Powers, and you reserved six. Mr. Quinn, you reserved three. That's correct. Okay, so let's start with Mr. Powers. Thank you, Your Honor. May it please the Court. I'd like to begin, if I may, with the claim construction issue on the 537 patent. The claim construction issue presented is whether the claim language that was added at the end of the very lengthy prosecution successfully in order to avoid prior art requires both the storage in the central database of configuration data, which the Commission found, and also the commands that produced that data, which the Commission did not find. And I would like to initially dispel two false premises, if I may. One argument that Cisco makes extensively in its briefing is that the issue below was either or, that Arista presented this question as the claim requires either configuration data or commands, but not both. That's false. We specifically said that the claims elsewhere require the storage of configuration data, so there's no dispute that configuration data is required to be stored in that central database. The question is whether the language that we're pointing to here on appeal also requires the storage of the commands. That is the issue. And so the Commission, I think, inappropriately followed the framing that Cisco had adopted and said, well, look at all the evidence in the specification that requires storage of the data. That's what we're going to find is required to be stored. That is not the dispute. Arista does not dispute that the claims require storage of configuration data. That's required by other claim language, explicitly. What the issue here is whether the commands also must be stored. There are three sources of support for that position. First is the claim language. Cisco argues that we're relying on highly technical rules of grammar, but they don't dispute that the correct reading of the claims using appropriate rules of grammar supports our position. They're wrong, though, in arguing that we're relying exclusively on... What's the rule of grammar? The proximity of the... Exactly. Exactly. And Strunk and White would say that if you wanted to read it as Cisco does, you would put in two different commas so that the modifier then is not modifying the intermediate clause, but the prior clause. And they poo-poo the argument as if we shouldn't be relying on the rules of grammar, but they don't dispute that the applicable rules of reading of the claim is appropriate. But we're not relying, as Cisco implies, simply on the correct reading of the rules of the claims. The specification and the file history also specifically support the position. Cisco's brief aggressively takes the position that the specification nowhere supports the idea that the commands are stored in the central database, are merely discussing the prior art. Not true. In the 537 patent itself at column 2, lines 51 to 53, that is explicitly talking about the invention, and explicitly says, quote, the sysdb receives configuration commands from various iOS subsystems, close quote. So Cisco's argument then devolves to, well, it doesn't say that that database stores the commands, it merely says it receives the commands. To which they have no response, though, to the argument that that's exactly what a database does, is store. And there's no other purpose for it to be sent to that database other than for storage. That's what databases do. And so Cisco's argument is then, I think, devolves down to something which is essentially nitpicking words, but not nitpicking the technology. The technology is, that database receives those commands, and that's what databases do, is store those commands. But not just 537 at column 2. The 752 patent, which was incorporated by reference specifically for the functioning of the database, says at column 3, line 36 to 38, the invention, again, not the prior art, receives configuration commands from a user of the router. The invention then communicates the configuration command to the centralized database system, close receives those commands, two places in the specification that are highly relevant, and no suggestion from Cisco that the database which stores things doesn't store them. And so the specification is, I think, quite clear. The prosecution history is, if anything, more clear. So the prosecution history, I'd like to walk through, if I may, in a little more detail than normal, because it's illustrative, and I think Cisco's description of it is misleading. The prosecution history is extensive. Year after year, Cisco tried to make, tried to get these claims allowed without adding the limitation that is now at issue. They then made that amendment, and they did so starting at appendix 1519. And they argued, in that amendment, that the key prior art reference called syscon does not teach router configuration data as stored in a database. So that's the argument that is part of this language that was added. And that argument was made by Cisco when this claim language was added. The Patent Office then did not accept that argument. They still maintain the rejection. That is at appendix 1537. And specifically, the Patent Office said at 1539, quote, it is noted that applicant has made no attempt to show why each and every amendment made to the claims have any sort of patentable novelty or advantage. Key, because that's the Patent Office saying to Cisco, you've added a lot of words, tell us why this makes a difference. Go back to why syscon makes a difference here. Does syscon teach the commands at all? Syscon, the argument was made that there was both a debate about whether it taught the commands, and a debate about whether it taught configuration data, and a debate about whether what was either or not equivalent to a command was stored in the database. Those are all three arguments that were made. The argument about the configuration data was not successful. They made it, and the examiner at 1541 said, acknowledged that syscon does not expressly disclose externally managing storing router data, but went on to say it discloses that databases for storing data are well-known, therefore I think it's obvious that conclusion is at 1542. Did the ITC find that syscon does not teach commands at all, whether they're stored or not? Syscon, there was a debate about that question, and the Commission, I don't think, went that far on that question. What the debate was in the file history, because we're not talking about... But if we find that that is the case, and that undermines your argument that you're making... No, because they made both arguments. They made both arguments, because the debate in the, we're talking about the file history now, the debate in the file history was, we don't think syscon teaches commands, or discloses commands, but they do have things called properties which maybe, which were being argued to be commands, and they said that's not stored in the database. And what's important here is what happens then is the Cisco's next response, and this is at 1564. At 1564, after unsuccessfully making the argument about configuration data, which is the argument they now have claimed won the day, they said, well, the examiner partially addressed our amendment by responding to the configuration data, but not the whole thing. And then says, syscon fails to disclose, teach, or otherwise suggest executing configuration commands before storing them in a database, close quote, them. So whatever... If it doesn't teach commands, then it doesn't matter whether they're stored or not. True, but they had both arguments, they were making both arguments as well, but the argument that we're talking about is not whether syscon really did or did not anticipate, or make, render obvious. What we're talking about is the understanding of the examiner and the applicant at the time about what the meaning of the limitation that was added meant. And Cisco was arguing that nobody would say that this language means you have to store commands. So whether syscon had commands or not, you cannot read this language from Cisco's patent attorney any way other than an admission that storing commands is required by the claims. Because he said, syscon fails to disclose, teach, or otherwise suggest executing configuration commands before storing them in a database. That is fatal to Cisco's current argument, which says the claim should not be construed to require storing commands in a database. After that argument was made, the claims were allowed. After that argument, the claims were allowed. And I think that that, for purposes of interpreting the claims, and there's obviously two roles that prosecution history plays in claim construction. One is interpreting the claims when there's a claim to ambiguity. And here, I think there is no room for ambiguity left in light of that statement by the examiner. Whatever syscon disclosed, because it was explicitly making the And the second role that prosecution history can play, which you don't need to reach here but is applicable here, is disclaimer. Even if it didn't require it, when you explicitly get around prior art successfully, you are not allowed then to say that that limitation doesn't apply. I don't want to let you go without asking a question off of what you're talking about, but I do want an answer. Cisco alleges that after it filed the complaint, you made certain business decisions, quote, to avoid the ITC's jurisdiction by allegedly no longer importing products that directly infringe. Does the record contain any evidence that showed you changed your business practice for some other reason? I don't believe the issue is relevant, but I'll obviously answer the question. It's not relevant to claim construction or any of the other questions. There was a debate about whether the importation of blank switches without the software was done for this purpose. There was evidence on that from both sides. I don't have, because that issue wasn't relevant to here, I don't have a record site for you, and it may not even be in your record because it's not an issue that's joined. That's just spice or poison that Cisco's trying to put in that is not relevant to any of the issues on appeal. So I know, and I didn't do the ITC case below, I know that was a disputed question. I cannot point you to the evidence on the other side. Let's move on to some of the other issues. What about the scope of the limited exclusion, the inclusion of components? Thank you. I'd very much like to turn to that. I think there's two questions that are very important for the court to address on that question. The first is a jurisdictional question, and the second issue is a question of the proper role, scope, and requirements for both contributory and inducement infringement allegations in the ITC. And both of those, I think, require a decision by the court. With regard to the jurisdiction question, I really think there's only two issues here. That there's, I don't think there's a meaningful debate that an act of infringement is required to invoke the exclusionary power of the ITC. That is clear from the statute and not, I think, seriously disputed. What is disputed, I think, here are two things. First, whether the commission actually did make findings with regard to infringement, the contributory and inducement infringement, from the components with a capital C. And our view, of course, is that the reason that the commission made reversible error, legal error, was it had an exclusion order as to components which had not been found to infringe. It specifically did not make a finding of infringement that importation of those components constituted contributory or inducement infringement. Why don't you challenge this issue in the other cases that, in the other patents that haven't been raised on appeal? There was explicit findings as to those and not as to here. So our point is those were two patents that are no longer relevant and not part of this appeal, but there were findings there. That's the whole point. As to the two PV land patents, the commission made explicit findings with regard to the components. As to the components on this patent, 537, the commission explicitly said it did not reach that question. Cisco tries to make a bootstrap argument that it sort of made a finding by saying it could reach the components based on the finding of infringement of the whole switch and the blank switch. That is not what it did. There is no way to reconcile the contrast between how the commission approached the PV land patents and the blank switches, as to which it made explicit findings of contributory. So these components are still excluded, right, under the other cases? No, because on the exclusion order, we've had proceedings in the customs office and everything else. We have a customs letter saying we can go ahead and import. There's no restriction at all. But the principle is important. So there's no restriction of the importation of the other patents? With regard to the prior ruling, no, because the functionality in the two other patents has been removed. And the Customs and Border Patrol decision was the removal of that makes it a non-infringing good. And that's a result of a customs ruling? Exactly. Exactly. And that I think is not an issue here either. So those can be imported for those purposes. So the first issue that Cisco raises, that there really was this sort of sub silencio finding, that doesn't stand. The second argument relies on VSCOFAN primarily, and says that the ITC has wide flexibility in fashioning its remedial orders. That as a general statement is true. But it does not have authority to exceed its jurisdiction. And VSCOFAN does not stand in the way of that. VSCOFAN is importantly distinguishable and materially distinguishable because the remedy that was at issue there was not the remedy for a patent infringement claim. Although there was a patent infringement claim there too. The remedy we're talking about from VSCOFAN was for a trade secret violation. And what VSCOFAN did was to say, there's a factory over in Spain, and they use the trade secrets. We can't tell here whether the trade secrets are being used or not. So it's not like a patent claim where you can assess it at the border and decide that is an infringement or not. And they looked and there's extensive debate and the issue was expressly joined in VSCOFAN, that said, the remedy for a trade secret violation is different from a remedy for a patent violation. The remedy for a trade secret violation is typically a complete ban on sale, or in this case importation, for a period that we decide would have been long enough that would have taken you to develop that technology without using the trade secrets. Okay. Let's move on to the next argument. Unless you want to keep going. I'm giving all the parties extra time on the argument here. I do not hear you. So I'm going to hold you to your time, okay? Thank you, Your Honor. All right. Thank you, Mr. Powers. Mr. Quinn. I don't know how I confused you with Mr. Powers. That's quite all right, Judge Reyna. Thank you, Judge Reyna. May I please the court, John O'Quinn on behalf of Cisco. Judge Wallach, I want to come back to the question that you asked a moment ago about, did they have any business reasoning for changing their importation practices? The commission at Appendix 518 found that they did not following the ALJ at Appendix 676. And that issue is directly relevant because it ultimately goes to the issue of remedy. In particular, it's relevant to the issue of remedy because it's relevant to whether or not there's a circumvention of the ITC's authority, both in terms of when the case was being tried and potentially a circumvention of the ITC's order here. And I think there are really two issues that are presented with respect to remedy. First is the question of, did the commission actually find that importing components, as opposed to the entire assembled switch, induce infringement? And I think when you look at the commission's opinion as a whole, you look at its adoption at Appendix 507 of the ALJ's findings, you look at the fact that it found that the exact same components induced infringement of the two VLAN patents, incorporating the exact same reasoning by reference in terms of state of mind and inducement, both with respect to the change in importation practices and with respect to the- So your opponent says there was no finding that the components infringed in this particular patent. Right. Were they silent about that or did the ITC actually make a decision and say they don't infringe in this case? The ITC did not make a finding to the contrary. The ITC did not disagree with the ALJ. The ITC, in fact, found that there was a corporate culture of copying, that there was induced infringement that was evidenced by the change in importation practices, and the commission adopted that with respect to the patent that we're talking about, the 537 patent, and it incorporated by reference that very same reasoning with respect to the two VLAN patents for which they have not appealed. And I'll also note that the commission, in its opinion, specifically distinguishes between capital B blank capital S switches and the switch hardware. Switch hardware being a term that was So I don't think you can fairly read the commission's opinion as not finding inducement with the components, but even if you did, then you would have before the court the question of the scope of the ITC's remedial authority. And the position that's being advocated by Arista raises, frankly, some of the very same concerns that you identified in your dissent in the original panel decision in Suprema, and I think that the en banc Suprema decision ultimately addresses, which is that you would commission into a paper tiger. If somebody can simply walk up to the border with a device, break it into two parts, take a step across, and then reassemble it. And that's exactly... Has customs issued an official ruling that's excluded the components from the exclusion order? So Judge Reyna, what Mr. Powers is referring to is that after the entry of the exclusion order, Arista has engaged in various purported design arounds with respect to both the VLAN patent and both VLAN patents and with respect to the 537 patent. There's currently an ongoing enforcement proceeding in front of the commission with respect to the 5-3, whether or not this is a successful design around with respect to the 5-3-7 patent. That issue is being litigated before the commission, but for the moment, Customs and Border Patrol have issued an order that is allowing them to import not just components, but assembled switches. And that issue with respect to the 5-3-7, it's not being contested on the VLAN patents, is currently before the ALJ in an enforcement proceeding. Now, I'd like to turn back to the issue of the interpretation of the 5-3-7. And Judge Reyna, I think you asked the key question here because the issue is whether or not CISCON taught commands at all. And that is exactly what the entire back and forth with the examiner was all about. And there's really a trilogy here in the prosecution history. It's all in volume three. It's at pages appendix 1521, 1539, and 1564 to 65. And when you walk through the back and forth between CISCO and the examiner, you see that the entire issue is whether or not the data that CISCO was claiming, data derived from executed user-supplied commands, was the same as the data that was derived from the structures at CISCON. And so if you look, for example, at 1521, the argument that was made is that CISCON does not teach router configuration data. This is, again, router configuration data as stored in a database. And further on 1521, the argument was that the router configuration data is data that results from a router configuration module executing configuration commands issued by a user. That's 1521 of volume three. This is the same time in which the claim language that we're talking about is introduced. And that claim language never changes. There's no subsequent amendment to the claim language. Now, the examiner does not appreciate CISCO's argument and doesn't understand the distinction that's being drawn between data derived from user commands and data from the structures that were at issue at CISCON. And that's the rejection that you see at 1539, where he says that the applicant argues that CISCON doesn't teach wherein the router configuration data is derived from configuration commands provided by the user. And he says CISCON does disclose this limitation. So that brings us to 1564-65 and the things that ARISTA relies so heavily on. And you have this prefatory paragraph on 1564. And yes, it does have the language that ARISTA quotes over and over that says CISCON fails to disclose, teach, or otherwise suggest executing configuration commands before storing them in a database. What is being said here is it's before storing the executed configuration commands. And I realize it's not very artfully written, but when you read the very next paragraph, it then begins specifically. And the prosecuting attorney then explains in the next three paragraphs what exactly was meant by this statement. And it's not at all about storing configuration commands. He says specifically, and then you look at 1565, he makes the point that the structures here, referring to CISCON, are not commands, i.e. user supplied commands. Makes the point that this is not, and this is a router. And then the bottom paragraph of 1565, finally there's no disclosure, teaching, or suggestion in CISCON that execution of user supplied configuration commands results in configuration data that is stored in a database. So it was always about whether or not the data derived from configuration commands was the same thing that the data was at issue at CISCON. And you don't have to take my word for it. The language is the same. It just seems to me that's pretty clear on 1564, CISCON fails to disclose or teach. It goes on executing configuration commands before storing them in a database. Sure. It's clear. I mean, the issue is there is when are they stored, before they're used or after they're used, but there's no doubt that they're stored. Well, it is, you know, this sort of gets a little bit to the ALJ's alternative finding, which the and that is whether or not even under Arista's construction, there would still be infringement, because the data is the executed configuration commands. So the configuration commands are executed. That results in data. The data is the executed configuration commands. And that is what is in fact stored. And to try to take this one sentence in isolation as Arista does and says, well, that must necessarily mean that it is the commands as opposed to the data that's being stored. That's exactly what this court recently said in MIT versus Shire that you can't do. That, you know, you have to look at this quote, selected sentence in the context of the entire response. And as this court said at 1120, 839 F3rd at 1120, quote, even if an isolated statement appears to disclaim subject matter, the prosecution history as a whole may demonstrate that the no clear and unmistakable disclaimer. And this court made a similar point recently in Trivascular versus Samuels 812 F3rd at 1065. And that's what you have here. When you look at this prosecution history, there's no disclaimer. This isn't there's no back and forth about commands. And this is ultimately confirmed by the notice of allowance at 1593, where the examiner now have a centralized database to externally manage router configuration data derived from configuration commands supplied by users. So it's all about those, the data being derived from the commands and that that was distinguished from the prior art. This is entirely consistent, of course, with the patent, the patent specification. Indeed, if you look at the, if you look at the title, the abstract, the summary of the invention, all of the examples, everything in this patent is about storing data, not commands, storing and then externally managing the data. And in fact, if you look at the first clause of Claim 19, element C, it talks about having a database that is that is configured to do what? To store data. Can you address the argument of your opponent concerning the usage of rule that we should apply here? Sure, Judge Arena. Well, first, I don't agree with with Mr. Powers that this is the his reading or Rister's reading is the more natural reading of this claim language in context. When you look at it in the overall context of the claim and you look at it in particular in the overall context of the specification, I think it is naturally referring to to the data in history as well. What's that? Mention the prosecution history as well. I agree. I think the prosecution history certainly drives home that this is all about the data that was derived from the commands. And of course, you know, Strunk and White may be about how one writes with style. Patent claims are often not written with style, and it's certainly not a I think the court can take judicial notice of that. But certainly it's not a tool for how you read a document that has already been written. And this court has made clear repeatedly, whether it's all the way back in the Ombach decision in Phillips or in the Finner case that we cite, that you look at claims against the best evidence and the best evidence is the specification. And if you look at column, we made this point in our brief, column 8 lines 46 to 53, the language that you find in column 8 is essentially exactly the same limitation that we're talking about here. And it makes clear it's the data that's stored. Now, in the time that I have remaining... You're right at your rebuttal time right now. Okay, well, in the time that I have remaining, let me make a few points about the 597 so that I have something to then say on rebuttal, because that is, of course, is our cross appeal. And first, let me be clear, I think the issue before the court is a question of law. There's no dispute about what the products do. It's just an issue of applying the claim construction to the undisputed facts. And the undisputed facts are that ARISTA has an element of its software that will stop, start, and restart agents based on certain information being received. The only question is whether it takes those appendix 14172. I'm going to walk you away from your undisputed fact because when a party challenges the IPC's application of the construction of the record before it, we review it as a question of fact. Well, I guess the point that I was making, Judge Wallach, is I don't think that there's really a dispute about how ARISTA's systems operate. I think the dispute is whether or not that rises to the level of being detecting. And in particular, if you look at appendix 14172, it's quoted at page 75 of our red brief. This is in the public record. It says, the process manager detects process exit and starts a new agent. And there's a similar point at appendix 14379, and it's part of the confidential record at page 74 of our red brief. But if you look at the sentence that's underlined there, you see, again, a demonstration about what they say their own systems do. The whole argument that the administrative law judge went off on, and which the commission then adopted, was the idea that reasoning by inference was not enough. And we submit it's certainly within the plain meaning of detect to infer something. They're not synonyms. Detecting by inferring is a subset, and we have some examples of how that works in our brief. But I would encourage the court to look at the examples found in figures 10 and 11 of the 597 patent. Because what you will see is that both examples involve concluding that a device is or a heartbeat. And you can see that at column 14, line 45, with respect to figure 10. Column 15, line 50, with respect to figure 11. This is exactly what we're talking about vis-a-vis ERISTA's system. And I think that if you look at the way that the 597 patent gives examples about what it means to detect, that certainly the limitation should be met here. I see my time has expired, and with the court's indulgence, we'll take any rebuttal time that you provide. Thank you. Thank you, Judge Ratner. Okay. Counselor Fischerow, you've got 15 minutes. Thank you, Your Honor. May it please the court. I would first like to address the component issue. ERISTA is an adjudicated infringer. We changed its importation practice with respect to its infringing switches to try to evade relief. The commission found that the switches infringed the 537 patent. The commission found that the importation to customers, regardless of the manner in which the switches are broken down for importation, are broken down for importation. The commission has broad discretion in fashioning the scope, form, and extent of its remedies, as long as the remedy has a reasonable relationship to the unlawful practices that have been found to exist. This is exactly what the commission did. Once the commission finds that an unfair trade act has occurred, it may issue an exclusion order under subsection D. Under Hyundai, VSCO fan, and sealed air, the commission has broad discretion in determining the scope of that remedy. Here, the components imported by ERISTA are used to make the infringing switch. So you're asking us to look at the ITC decision as a whole, and they found that the components were infringing in several other patents, but not in this particular patent, or at least it's silent to this. This is what your opponent is arguing. Your Honor, I submit that the commission found that the components do induce infringement. In this particular patent? In this particular patent. Where in the record is that? I would say that would be at the commission's APPX 528. I think it's encompassed within the commission's analysis on this page. The commission here talks about the switch hardware. Earlier in its opinion, it also refers to the switch hardware at APPX 511. Go back to 528. Where is that? The commission talks about the accused products and the blank switches, and then it analyzes the switch hardware. Here, switch hardware has been defined to include the components. If you look at the context from APPX 511, there the commission's use of switch hardware, it's clear that it refers to the components. But regardless of whether or not the commission found that these specific components induce infringement, the commission's remedial authority to stop circumvention of the commission's order here, when this case began, circumvention is a different issue. But what I'm interested in is, are you arguing that the ITC can make a blanket assertion that all of the components, with respect to the patents it's reviewing, that they're all infringing, and they're all subject to limited exclusion order? Here, or rather, let me take that back. I'm sorry. That the components are found to be under these patents, but it makes exileness to the one we're reviewing now. And by that, we are to extend that finding to the patent in which the ITC has not made an explicit finding, and find that the limited exclusion order covers that. So if you don't accept the commission's position that APPX 520... The appendix, excuse me? Yes. Yeah. The page we were just talking about, about the commission's opinion, makes explicit findings on components. Components are included within the commission's finding on the switches themselves. The commission found that the switches infringe, the blank switches infringe. And in order to prevent circumvention, the commission did not make, with respect to contributory infringement, explicit findings as to the components. Because ARISTA had, during the course of the investigation, had a pattern of conduct of changing what it was importing. And it leads to the question of, if we adopt ARISTA's approach, every time a respondent changes their importation practice, whether they take off one component, import it, and then we have to make a new finding. Here, the commission saw ARISTA's pattern of conduct and determined that it made a finding of infringement as to the switch and the blank switches, and it didn't need to reach the individual components. For the future, I would like to see the, if that's the case, and I'd like to see the commission and its decisions making those type of blanket-type findings. Otherwise, I'm going to look for a finding that's explicit to the patents under review. I understand what you're saying about the circumvention, and that the circumvention aspect may have been applicable to all of the components of all the patents. But I need to see that. I'm hesitant to extend findings made specifically for other patents to every single patent that's I mean, go on with the argument. I'm just saying, telling you that to take back home. Yes, Your Honor. I think, though, if you look at APPX 527, where the commission expressly discusses the contributory infringement, it specifically says, if the commission, if ARISTA attempts to circumvent commission remedy by importing only the components of the accused products for reassembly into complete functional switches, it would still be in violation because it finds the blank switches and the fully assembled complete switches indirectly infringe. So here, the commission is saying, we found infringement, and to avoid the circumvention, we're not reaching this because ARISTA is going to continue to change its importation practice. And under Hyundai, VSCOFAN, and sealed air, the commission has broad discretion to do this. As this court stated in VSCOFAN, the FTC's wide discretion in fashioning the scope of its remedy is applicable to the commission. And in Rubberoid, the FTC specifically stated that the FTC's legal practice in the precise form in which it is found to have existed in the past, and that the commission must be allowed effectively to close all roads to the prohibited goals so that its order may not be bypassed with impunity. This is exactly the type of circumstance we have here, that the commission was seeking to prevent circumvention. Could you address the claim language of the claims is clear. Even if the court doesn't agree with that, the case law is clear that you look to the context of the claims. Here, that is the specification, the other limitations, the other claims, as well as the prosecution history. And nowhere does this specification teach storage of commands. Yes, as Mr. Power stated, it teaches that data is sent to a database, but at no point does it talk about it being stored. Instead, as Mr. O'Quinn pointed you to, column eight of the specification, that teaches that it's the router configuration data that is stored. Turning to the prosecution history and the office action at issue, I would like to walk the court through this office action just like Mr. Powers. And if we start at APPX 1564, in this response, the applicant explains to the examiner that SISCON does not disclose, teach, or suggest the execution of user-supplied configuration commands that results in configuration of data that is stored in a database. To arrive at this point, the applicant directly responded to the examiner's rejection and used the term specifically as it addressed the columns and lines that the examiner relied on to reject these claims. And the applicant at the top of 1565 explains that there are no commands in SISCON. After that, in the very next paragraph, it explains that SISCON and the execution of the comparison of the data fields in SISCON is not the equivalent of executing configuration commands. It is simply a comparison of the data structures. And then, finally, in that last paragraph, the user talks about what is stored. And just like the specification, there's no real discussion of storage other than this paragraph and the one sentence taken out of context that ARISTA relies on. And in this sentence, the applicant expressly says there is no disclosure, teaching, or suggestion in SISCON that the execution of user-supplied configuration commands results in configuration data that is stored in a database. That is talking about what is stored here. And the them in the sentence at issue that ARISTA relies on must be the configuration data. That's what the rest of this office section is talking about. It's how that configuration data is derived and the storage of that data. The commission submits that the claim specification and prosecution history are clear that what is stored in the database is the router configuration data. If there are no further questions on this issue, I would like to turn to SISCO's appeal. Substantial evidence supports the commission's finding that the asserted three mechanisms of ARISTA's products do not detect a change in configuration as construed by the commission. This means that the evidence is sufficient that a reasonable mind might accept the conclusion. Here, that is the case. SISCO asserts that the commission's finding actually supports position, but this is not so. The commission found that the first and second mechanisms at most infer whether an agent is operating by relying on the testimony of Dr. Duda and two other documents. I'm not going to go into the evidence that the commission relied on because it's on the confidential record, Your Honor, but this evidence supports the commission's determination that this is not a change in configuration as defined by, as construed by the commission. And turning to the other two mechanisms at issue, I would like to point out most of this is also on the confidential record, so I'm going to leave it to Mr. Powers to discuss this. I was going to talk about mechanism, the third mechanism that, and I'll turn to Estapo, Your Honor. This third mechanism also is not detecting a change in configuration, but instead is detecting something else. And the evidence that the commission relies on, which is testimony that is found at APPX 11288 through page 9, and then APPX 6018, and that's Dr. Hollingsworth, and then testimony of Dr. Wicker at APPX 15005 through 06, 18999 and 19000, as well as 6068, all demonstrates that this third mechanism at issue is not detecting a change in configuration of the subsystem, but instead is detecting a change in configuration of something else. Here, CISCO takes snippets of a document the commission relied upon out of context. If you look at the entire context of this document, which portions of it on APPX 14378 and the top of 14379 describe exactly what is happening here. And Dr. Hollingsworth's testimony at 11288 through 9 is in line with the commission's interpretation of this document. On the issue of prosecution history or estoppel, if the court has no questions on the infringement issue, the commission adopted the ALJ's finding that a signer estoppel prevents ARISTA from challenging whether or not the 597 patent is directed to patent in an eligible subject matter. Here, the inventor of the 597 patent at CISCO, and later a founder of ARISTA, stated in his assignment that he is, quote, he is the original first and sole inventor of subject matter, parentheses, processing, machine, manufacturer, or composition of matter, or an improvement thereof, which is claimed and for which a patent is, which is sought, sorry, and that can be found at APPX 14249. This assignment didn't include a carve out for changes in the law. This court in CLS blank said that 101 is not a threshold issue. So here, the commission submits that it appropriately determined that a signer estoppel applied in this circumstance. So no further questions, Your Honor. Thank you. Mr. Powers, I made a mistake. You'd actually run out of time, but I'm going to restore you to three minutes, okay? I'd like to begin with the claim construction issue from A. The question before the court is whether the language that read naturally requires storing of commands is sufficiently ambiguous not to require that. The file history, I think, eliminates any possible ambiguity. The statement we rely upon is simply not understandable in any other way than being consistent with the requirement that the commands themselves are stored. No matter what SISCON had, they said it, they're stuck with it. It's not a question of disclaimer. It's a question of understanding the meaning of the words they used and watching how they use those words in the file history. Counsel for SISCO argues that, well, it may not have been artfully written. Well, we think it's pretty clearly written, and the argument in the file history is quite clear and unmistakable. Counsel for SISCO argues that at 1565, it's all about the data, 1565 of the appendix. The very next sentence, though, and he was fairly aggressive in saying after that snippet, it's total all about the data. Not so. The very next sentence on the portion he follows, the novelty here is that this claim limitation provides a way to incorporate a database into managing user-supplied configuration commands. That's the very next sentence. It's not all about the data. It is partially about the data, for sure, but not all about the data. He then relies upon the notice of allowance to say, well, this argument didn't make any difference. Again, not so. When you look at the actual notice of allowance at 1593, it says, in addition, we rely on applicants' arguments, which include the argument that ultimately won the day, in addition to the rest of the claim limitations. So the arguments that are made simply cannot be reconciled with the statements that their patent attorney made. They then argue, well, that's merely a snippet out of context. Again, not so. It's the culmination. They tried to get the claims allowed by using the data limitation. That wasn't enough. The made this argument, and it was enough. It's not a snippet out of context. It is the thing which got it allowed, because that is the only new argument that was made. Now with regard to the ITC jurisdiction, this is not a question about circumvention. Not. Because the issue isn't whether you can change something and get in the door otherwise. The question is a question of jurisdiction. Had the ITC made a finding in this case as to this patent, we wouldn't be here on this issue. And if they made a finding about that, they have various ways of avoiding circumvention of their orders. The point is they did not make that finding. They specifically said we don't need to reach it after going through great detail and finding in contributory and inducement infringement. Where did they say that? They said that at 527 and 528 of the record. And you compare that with 550 and 551, where they made explicit findings on other patents. There is not, there's simply not a way to find an inferred finding on this record. But if you're dealing with the same components for all of the patents, then what difference does it make if they make an explicit finding for three other patents but not the fourth one? They're different inventions. It's for contributory. It's the same component. Because for contributory infringement it has to be a material part of the invention. The PV land patents are a different invention. The fact that it's material to PV land doesn't mean it's material to this. For inducement, the requirement is that it actually induce infringement. The fact that it may have induced PV land doesn't mean it induces this infringement. They are different inventions and you must prove it and find it. That's necessary for their jurisdiction. As to 597, the issue is not a question of law. It's specifically a substantial evidence in the record. The commission's finding as cast for the ITC, that finding was extensive. And what Cisco's argument boils down to is you can infer whether a process has stopped running. But that is not a change of configuration or state. Under the claim construction the commission used, which Cisco has not challenged, that commission finds the configuration changes are limited to those in the 597 patent. Not all changes. And it enumerates those. Okay. Mr. Prosser, I'm going to ask you to conclude. Thank you, Your Honor. The only other issue I wanted to reach was the Estoppel question. There's no representation in the inventor's oath about the legal question of whether that's eligible subject matter. You can read that oath ten times and it won't be there. So Estoppel requires a Thank you very much. Mr. Quinn. And you have three minutes. Thank you again, Judge Rayna. I'm happy to answer any questions the court may have, but otherwise I'll just address the 597 issues on rebuttal. First, just so it's clear to the court, we and the commission are in agreement that a sign or Estoppel does apply to the challenge of invalidity. The disagreement between us and the commission, of course, is whether or not there's a 597 patent. And a point that Mr. Powers made, I think, is an important one. If you look at Appendix 537, the commission found that the change in state of certain things is equivalent to a change in configuration. And there's no question, there's really no dispute that there is a change in state with respect to how ERISA systems operate. And again, if you look at the evidence at Appendix 10336 to 338, there's a combination of certain features that work together to directly detect that an agent is no longer running. That is a change in state. Similarly, if you look at Appendix 14172, and some of this is confidential, so I won't discuss the details of that, but they all confirm our position. And indeed, the whole argument that ERISA made to the ALJ, which the ALJ did not adopt, was that there was no change in configuration, that a change in state was not the same as a change in configuration. The ALJ then, for the first time, adopts this approach to detecting, saying that detecting doesn't include inferring. But of course, the plain meaning of detecting does include the term inferring. And to the extent there's any doubt about that, and you heard what Mr. Powers said about reasoning by inference or by assumptions not being enough, that's belied by the examples that are found at Figure 10 and Figure 11 of the 597 patent. And ultimately, it's really no surprise, or shouldn't be a surprise, that ERISA switches would have these patented features, given the ITC's finding that ERISA has a corporate culture of copying Cisco, as well as the fact that this technology was developed by ERISA's founder while he was a Cisco employee and assigned to Cisco, and that is exactly why, of course, if the Court were to reach the issue of a sign or estoppel, a sign or estoppel absolutely should apply, and there's no basis for a Section 101 exception to a sign or estoppel. If the Court has any further questions, I'm happy to answer those as well. Otherwise... Are the components that we're talking about here, it's a switch hardware, correct? That's exactly right. It's components, and you can see them explained by the ALJ at Appendix 682. It's processors, memory, CPU cards, chassis, switch cards, and fan modules. Okay. No other questions? We thank you, Mr. Poon. Thank you, Your Honor. This Court's now in recess. Let me just tell you, Mr. Oakman, I disagree with you about the lack of style in patent writing. I think it's baroque.