Oral argument — MICROSOFT CORPORATION v. DATATERN

The first case this morning is Microsoft v. DATATERN. Mr. Belk? Thank you, Your Honors. Good morning. I'm Eric Belk. I'm from Carter & English in our Boston office, and I represent DATATERN. The district court took the claim out of claim construction. The claim wording, as we know, is the starting point for a claim construction. We're talking about the 402 patent? That's what I'm going to talk about primarily, Your Honors. What the district court did in the 402 patent, 502 as well, is to jump ahead and to start the process with examples in the specification, or the prosecution history, or in some cases, extrinsic evidence, while bypassing the claims. And the result of that, Your Honors, is that the district court rewrote claim one of each patent by inserting new steps into each of the methods, both patent or method claims. So let me repeat that. The district court, in effect, took out a pen and inserted entirely new steps into the claims. These steps are not present in the claim wording, and are not otherwise required by the specification or the prosecution history. The district court also erred in accepting subject matter jurisdiction in the first place, and I will speak to that if I have sufficient time. Well, let me ask you about what I think is your first point. You're not disputing that the logical tables require normalized data, right? You're just disputing whether or not that normalization process is required in a particular claim. Right. Am I right about that? I don't think they even require normalized data. The logical table is simply a representation or a definition that exists virtually. That's what logical means, is to the user, it appears as if the user is working with the real tables in the physical database. But really what the logical tables are is, if I could back up, as you've seen from our briefs, there's a mismatch. The relational database is structured in one way, through tables. The object-oriented application has a different structure to it, so they don't necessarily readily speak to one another. What a logical table does, Your Honors, is enable... Do I understand your argument to be, among other things, that there are at least two disclosed embodiments, that in Figure 1 and Figure 5, that don't require any normalization? That's right, Your Honor. So, Figure 1 reads the relational schema, which is the structure of the database, directly into the process and creates relational schema objects. That's why they're called relational schema objects. They are objects that are representing the logical... I'm sorry... That is normalized before it gets there, right? It could be normalized before it gets there. In fact, what the patent says is, typically, the relational database is organized into already normal tables. If you look at the background of the invention, Your Honor, in Column 1, right there it says, starting in File Line 2526, it talks about relational databases. They are typically organized into... they use normalizations to minimize duplication, and so you have, typically, in a database, already normal tables. They can be directly read in through the capture process to create the normalized relational schema object. That's what a normalized relational schema object is. Now, is it your argument that Figure 1 requires no logical table? No, that's not our argument. The claims require logical tables. That I understand. Where is that in Figure 1? They're not in Figure 1 or in Figure 5, for that matter, Your Honor. Figure 5 is the embodiment that Microsoft relied upon. Then is this embodiment in Figure 1 claimed, since it has no logical table? Of course it's claimed, Your Honor. The normalized relational schema object, according to the claim, represents the logical table. One way to do it would be to read in the normalized relational schema objects, create those, you form those, and you define logical tables by selecting columns that you need. You may not need all of the columns in a particular table. Does this mention of RSO may be helpful for you? Doesn't it mention RSO as opposed to NRSO? It does, and that means... Is that part of what I understood your argument to be? I'm just trying to make sure I get it. It's quite complicated. It is complicated, Your Honor. I think it uses the term RSO because in the database you could have either tables that are already normalized or you could have denormalized tables. For example, Claim 8 is a claim that discusses starting from denormalized tables and some of the later claims as well. The reference was made to both Figure 1 and Figure 5. In Figure 5, and this is an argument I think Microsoft makes, it's described in Column 4 of the patent. It talks about the normalization process allowing the user to define different tables. So it seems to me, and the normalization process does appear in Figure 5, so how does that help you? I'm not understanding how Figure 5 helps you. It wasn't our argument. It was Microsoft's argument. Figure 5 is simply an alternative embodiment. So that does include the normalization process. Figure 5 is a different column. So that supports Microsoft's argument. Your only argument is that that's not required in all instances, right? That's right. Claim 1 is broad enough to support either claim, either Figure 1 or Figure 5. You didn't correct me, which you should have. I said Figure 1 and Figure 5 both do not disclose normalization, and you said yes. You should have corrected me, because Figure 1 doesn't disclose normalization, but Figure 5 clearly does. Don't let me misspeak, even when it helps you, because trust me, they'll figure it out. I'm sorry. Maybe I misheard you, Your Honor. So Figure 5, and that was our argument, so there are at least two embodiments that are disclosed. Figure 5 shows the alternative normalization process. And you're right, that is discussed in, for example, Column 4 and Column 10. And so what the district court did was, there are claims that specifically call out a normalization step. That is not in Claim 1. Claim 1 wouldn't necessarily exclude a normalization step, but it's certainly not required. And in addition to inserting into the claim, in essence, this normalization step, what the district court did was, we did not just any normalization step, but the normalization step that has to do with normalizing a database. That is the normalization that is discussed in the background of the invention, and the district court specifically refers to that. And the background normalization is concerned with how do you structure a database so that the data is organized efficiently to minimize duplication. Even if we would agree with you on the claim construction, where does that leave us in terms of this case? Just send it back for remand, and you still have to determine infringement based on a revised claim construction? Well, we actually have, in the summary judgment briefing below, we submitted testimony, much of which was not opposed. So you could actually reverse. We're talking about going what? Oh, the infringement and the way that infringement would work if the claim construction was reversed. But certainly you could remand if the claim construction were reversed for further proceedings. And there would still be a dispute, arguably, between whether or not there's infringement. I think that would certainly be Microsoft's position, yes, Your Honor. On the 502 patent, if I understand your reply brief, you believe that we should deem the arguments all waived in this particular appeal because of the attempt to incorporate by reference, is that correct? That's right, Your Honor. You believe that incorporating by reference is a violation of Rule 28I under these circumstances? That's right. That's a pretty draconian response that you're asking us to take. Do you want to defend it a little bit? Well, first of all, the rule is the rule. And I think what should have happened was Microsoft should have at least moved for permission to do so and should have pointed out what are the similarities and what are the differences. Now, there are certainly differences between SAP's arguments and Microsoft's arguments, Your Honor. But Microsoft's position is to affirm the district court. I mean, even if we were to exclude the arguments made in the other brief with respect to this case in 502, we've still got the district court's opinion. Right, that's right. And we have our arguments as to why the district court erred with respect to the 502 patent. So, yes, I do agree that you still need to conclude that our arguments are correct. Well, and of course, since they made the argument for the 502 in SAP, they are fairly presented there. So if we just decide the SAP case a minute before we decide the Microsoft case, right? And I understand that, yes, you're absolutely right, Your Honor, and this is something that we discussed. There is the practical implication of even if the rule is violated. That is clear. There are differences, however. And one of the differences, in addition to, for example, different products. SAP has its business object products. Microsoft has its own products. And we conceded infringement differently. So some of the claim terms, one being in the 502 patent, the creation of the interface object, that step. We conceded as to SAP on one portion of the reading of that claim. It conceded as to Microsoft as to a different portion of that claim. So there are some differences, and we do point that out in our briefs, Your Honor. So to the extent that Microsoft seeks to rely on SAP's arguments, they don't quite mesh up all the way. Now, Mr. Bell, let's back to normalization. Assuming that normalization is not required by the claims, what happens when we start with denormalized physical tables? So, for example, if you look at Claim 8, you have a process where you start with a denormalized table, and that gives you a process for doing that. And what has to happen? There has to be normalization, right? I'm not actually sure that that is necessarily so. Certainly there is an example. How could it not be? Maybe that's what you could tell me. So you are defining a logical table, which is simply the step of first selecting one or more columns from the table, and then you are assigning a primary case. I guess you would need to have normalization. I think you'd have to have normalization, if I understand it correctly. So really, the only time normalization is not necessary is when the tables are already normalized. Wouldn't a person with skill in the art kind of automatically understand that? I mean, not even automatically. It's almost beyond presumption, isn't it? I think that's right, Your Honor. So normalization, as the district court says, is necessary every time it is necessary. That was somehow circular, but I think you understand what I'm saying. It's necessary in the claim every time it's actually necessary in practice, isn't it? I think that would be right, Your Honor. So is the district court correct in its claim construction? Because the only time when normalization isn't there, it really isn't necessary anyway. Because you require it every time it's going to be put into use. And a person with skill in the art is going to understand the distinction, and so the district court kind of made this simple, right? No, I think it actually makes it more confusing, Your Honor, because the district court is saying a normalization step is required. Even in those instances where you don't need normalization, for example, when you're starting with normalization. When you've got normalized tables, that's the only time. You don't need it. But wouldn't a person of skill absolutely understand that? It's already normalized. I don't need to re-normalize. Well, how is a jury going to know that, Your Honor? I think I just explained it. If the district court says a normalization process is required, then a jury is going to conclude that it's going to be required no matter what the table is. And that's just not right, Your Honor. And not only did the district court read in a normalization step, but the wrong normalization step. The normalization step from the prior art, I should say, of structuring the database to minimize duplication by dividing the tables. That's not what the... And we're just talking about claim one because the other claims are all covered. That's right. Other claims specifically require a normalization step. Claim one does not. It's broad enough to cover both embodiments. Thank you. I think we ate up your time. We'll give Mr. Rinas an extra two minutes and then you can have your rebuttal time back. Thank you very much. We'll give Mr. Rinas two more minutes. There we go. Time will be even now. Thank you. You may proceed. Your Honor, may it please the court, I'm prepared to address normalization because it's essentially the subject of that argument. But I'm happy to answer any questions about anything. I would like you to start with the incorporation by reference and what we ought to do about the incorporation by reference in this case. It is not consistent, as I read, both 28I. Don Dunner out there in the audience writes lovely briefs, but you don't just get to incorporate them by reference whenever you file your own brief. That's correct, Your Honor. So, first of all, I hope you appreciate the purpose for what we did without consideration. Really? Because I think the purpose for what you did was to extend your word count. Mr. Rinas, how many words over 14,000 did your brief go when you incorporated by reference? 3,025 words. So those are 3,025 words that you are not entitled to in your brief if you would type them in. So I think your purpose was to get more words to mount a bigger defense. Okay. I don't believe that's so. I could have split the argument and made most of the argument on some terms in one and most in the other and gotten more pages because we didn't use the last 15 pages in one of the briefs. But that brief didn't raise two patents. This one did. You needed more pages in Microsoft because you needed to respond to two different patents. We could have gotten more pages if we had split the argument into the two briefs. We wanted to do – it was out of consideration to put the argument in one place because it's the same argument about the same order. Let me argue on the law on your assumption that it's a violation because I don't believe that's right either. First of all, if you think about it, we could have gotten more if we had split it, but what if I didn't consider it? Second of all, Rule 28 is permissive. It doesn't preclude, and it notes where cases are consolidated. You can incorporate it by reference. This opinion is the result of consolidated proceeding. It's not just that the cases were proceeding below. It's one opinion. It would make no sense to argue about one claim construction order, which is an issue of law, in two different briefs and split it out. Well, it was consolidated initially, right, for purposes of pretrial, but then not – I mean what was the consolidation? So what – this is – I mean if we're using our time on this, this is a result of the fact that there was no motion for a consolidated appeal. I didn't know – we didn't know what they were going to argue about which. Issues came in and out. There's a lot more issues that aren't being appealed here. So I had to wait to see what their briefs were. They split the argument in their two briefs. I was confronted with do I move to consolidate, maybe only get one brief, create a little bit of a rush, or do I do what I did, which is not use the last 15 pages of one brief and put it in one coherent place. But on the law, Rule 28 permits it in consolidated cases. This is clearly consolidated proceeding below. The opinion below was itself consolidated. Also, I was keeping in mind the practice note to Rule 28, which advises when multiple parties are represented by the same counsel that it's inappropriate to split the argument. But you recognize, of course, this is a federal rule of appellate procedure, and they're talking about consolidated cases, and in the same sentence talking about the appellants and the appellees, clearly they're talking about consolidated on appeal. Not sort of consolidated at some time in the past, but no longer consolidated cases. I think the opinion below that's on review is still consolidated. And I also think that there's the practical issue that it's an issue of law that's being decided. But I think most importantly it's about motives of counsel. I think if you think about it, you'd understand how I could have gotten more pages if I had split the argument across two briefs. On the merits. It's not in the claims. It is, actually. It's in normalized relational scheme object in one of the claim terms. The word is used four, five, six times, but normalization, the process of normalization is not. So I think the important thing to understand is in the patent itself, in the specification, there's no reference to NRSL normalized relational scheme object. That concept first came in to the patent in an amendment to overcome prior art. And normalized got tacked on and added. And I think if something's a normalized object, that means it's been normalized. I mean, it's the past participle. That's what it means. I don't see how one would assume that the object's not been normalized when that's exactly in the claim term. In terms of the logical table, there's a specific definition. So I think what is a logical table? We've dealt with logical tables in a variety of different technologies and it's different meaning in different places. There's a definition. I believe it's a definitional quality on column four at line 28 through 34. It says once a denormalized table such as table AP is captured from the database, the normalization process allows the user to define different tables using subsets of columns in the table AP. And that's figure five. Those references are all figure five. Clearly no one's disputing, including your friend on the other side, that that's the normalization process. But this is where logical table is first introduced. It says this new table is hence a subset of a real physical table in the database and is henceforth referred to as a logical table. I believe that's definitional. I believe that's first introduced at this point. And I believe that this reinforces the claim language of the normalized relational object. It got normalized. It's just plain grammar. So I think normalization is clearly part of the claim requirement and that's why all the other claims claim it in different ways. It's just not surprising that the special sauce here is the logical tables. I think what resonated with me was Judge Prost's point at first. It's sort of hard to deny normalization is part of it. I think that's right based on the arguments I just tendered. The question is this bit about, well, what does normalization entail? And I think normalization clearly entails the dividing of the two tables within a logical table construct. And the reason that I do is for several reasons. One is that's what the term means conventionally. Council relied even on the reference at column one at lines 29 through 32. Where it puts quotes and defines normalization as dividing the data into different tables. But it's also reinforced, and I think this is... Wait, what was your reference just then? Column one at line 30 to 32. Okay, but I think to me the real sort of... At that location, does they define logic table as a subset of real physical table in the database? I think that's a descriptor of it. I consider the definition to be in column four. I think that's more helpful as such. But clearly that's an element. I mean, when you divide... I mean, the point being, Your Honor, that when you divide into two, it's a subset. If you divide one thing, it's multiple. Each one's a subset. So it's totally consistent. But to me, the trick to getting this, to understanding our argument on normalization, what it is, when you accept that it's got to be part of it, for the reasons I set forth at the outset. So you look in column four, and in this definitional phase, it says at line 30 to 32, it says allows the user to define different tables using subsets of columns. That's the reference that Judge Moore made. Of the table AP. You see that? Of the table AP. What line are you on? That's at 32. 32. Column four, 32. And if you look at figure seven, which is very, very important, this is your normalization process. This is demonstrating within the logical table construct how you would normalize. And you'll see the references to table PL, table AL, and table AP. At least if you have that. The table AP is the denormalized, and then table PL and table AL are the dividing out, just like standard normalization, I don't know how, it's a complicated stuff, but at least if you study up on this art, there's a basic construct of what normalization is, taking a superset and dividing it into subsets. And that's exactly what's shown here. So the idea that it's a different kind of normalization is completely inconsistent with figure seven, which... So this is only discussing an embodiment which is in denormalized form. Yes, if you... The tables are in denormalized form in this... That's correct. I'm starting, this argument is starting from the assumption that normalization is required by the presence of logical tables. I think your opponent agrees. When it's denormalized, you've got to normalize it. What about the other side of the equation? If it's normal, then there's no normalization and it doesn't need to be normalized. I'm putting... You're saying that's the part that will confuse. I think that when you have normalized relational schema objects, they've been normalized. That's what it's all about. They had to add that to the claim, and that if it's just a relational schema object, without having been normalized, there's no logical table. Just think about it, Your Honor. Okay, but here's the problem. These are method claims, okay? And if the method claim requires normalization, the fact that no normalization is required creates a problem, right? I mean, I guess I'm just finding it difficult to reconcile your answer to Judge Rader's question, because if there are circumstances in which this method is not required to perform normalization because it's already been done, there's no need, then... Then it's not using the method as a claim for our argument. Let me back to persuading you that normalization has to happen. To me, the second debate is more interesting. On that issue, what is critical to understand is this long discussion of Figure 1 that's been mentioned, and they go into it, there's no reference to logical table. Logical tables are the most important thing. I think we can all agree on that. Summary of the invention, abstract, everywhere, right? They're the most important thing. How could you have an embodiment described in Figure 1 that's talking about the invention itself that doesn't even introduce the concept of logical tables until four columns later? It doesn't make any sense. That's table setting. What they ended up focusing on was the logical table, which is introduced for the first time and defined in Column 4, related to Figure 5. I'd like to ask Mr. Reines a few questions on DJ jurisdiction, but I don't want to make him bounce back and forth. Let's go there. I want to go there, too. Mr. Reines, can you explain to me the basis, as you understand it, for why there is DJ jurisdiction in this case? Yes. Under IRIS, there's two problems. We can go even higher up. MetaView requires that there's something touching a legal interest of the party. In IRIS, the way that was implemented... Well, not something touching. There has to be an immediate and real adverse legal interest. In MetaView, they use the word touching, but it has to be... I think it has to be immediate and real adverse legal interest. The dispute certainly has to be immediate and real, so I have no problem with that. IRIS implemented that with a two-pronged test, as I read it. One is that there's a claim that could have been brought, which is also consistent with microchip, which is pre-metaView. The second thing is that there be an indemnity claim. On the could have been brought, clearly the claim could have been brought. Why? Why could it have been brought? What is the claim that you think could have been brought? Indirect infringement. Exactly what? Because that's just a label. 271B or C. Both. Both. You think a claim against Microsoft on this record legitimately could have been brought. Do you think that if there's absolutely no evidence anywhere in any record for anybody of intent for inducement, for example, that it would be fair to say a claim for intent could be brought? I mean, that's a self-answering question. No, that wouldn't be. But since the claim is brought, and the judicature... No, no, no. Those are post-complaint acts. Our case law is crystal clear that post-complaint acts cannot be considered for determining whether jurisdiction exists. So you're talking about the counterclaim, which was compulsory. No, what I'm explaining to you is if there's a question whether something could have happened at time A and it does happen at time B, that is probative of the fact that it could have happened. No, no, no. We've addressed this exact fact pattern and quite clearly said that post-complaint behavior cannot be considered for purposes of determining whether jurisdiction exists. They had to file it. It was a compulsory counterclaim. They didn't file it. They lost it. They didn't have a choice. Well, you wouldn't lose something you didn't own. That logic is not compelling to me personally because if you didn't have a counterclaim, you wouldn't bring it. I'd like to know what was the basis upon which you said both B and C. And so what Eris did is said contributory infringement was the basis. And it actually walked through all the elements and showed how the infringer had, or potential infringer, had a legitimate concern about adverse legal interest on every single element for contributory infringement. I see no similar thing here. It seems to me you're alleging nothing more than they sued our customers on our product using claim charge, referring to addo.net. Therefore, we could potentially be sued for inducement or contributory. And that doesn't seem to meet what I understand is the requirement. Everything Eris has, this case has, except the participation. Well, it has all sides of that, the participation and the licensing. Everything Eris has, really? Because in Eris, there was, as I understand it, let's see, let's look at Eris. All through the claim charge is reference to ADO.net. The name of the claim charge is customer of ADO.net. They're not even customized to the different customers. It's so clear that what's being accused is the product under a contributory theory, I think, and then an inducement theory. But really, where is the evidence in those claim charge or otherwise that it was that ADO.net was especially made or adapted for use to infringe the patent? Because I don't see that evidence anywhere, but that existed in Eris quite clearly. Judge Dykes, in his majority opinion, explained it. So where is that evidence in these claim charge? I think it's just throughout the claim charge by saying that there's a normalizing. But ADO.net, Mr. Reines, has lots of uses beyond simply using to create an interface that will allow for object-oriented to communicate with relational databases, correct? ADO.net can be used to retrieve data broadly in many sets of circumstances, can't it? If there's an NRSL functionality, that functionality would only have one use, right? That's the argument. If you're saying that software, because it can have many uses, can never do that, then in the case where someone sued 60 customers in one jurisdiction, you can never come in if they're trying to avoid keeping the manufacturer in, because it's software and therefore it can always be used in different ways, especially in a massive operating system or whatever else, then yes, then maybe there's never jurisdiction. I don't think that's why it's policy-wise. I guess here's my problem. My problem is that I don't see you having any evidence in the record that Microsoft has a legally adverse interest because the inducement standard requires a very high level of intent, and I see no such evidence of intent. And then for contributory, there's all kinds of elements that aren't established. And in the heiress' opinion, Judge Dyke shows how the record in that case gave the alleged infringer a reason to believe that he was in jeopardy on every one of those elements, and no similar thing happened here. I didn't see anything in heiress' that there was any show of knowledge as a patent, those kind of things in there. Well, actually, it is present. And the same kind of intent. No, I mean heiress' is not an inducement case. Heiress' is only a contributory infringement case. Yeah, but I think those standards, if you read, the way I read GlobalTAC is that that applies to both. Wait, you think the same intent standard that we articulated in DSU for inducement applies to contributory? Well, DSU was overruled by GlobalTAC. I mean, that was a different standard. But you think the same inducement standard of intent in Judge Dyke's recent appeal decision applies for contributory? There's a good argument, yeah. But in heiress' what they said is you obviously don't have to admit that you're inducing infringement or that you're contributing infringement in order to get your guy in court. The whole point is you can contest that. But I don't see any basis upon which you ought to be shaking in your boots. I mean, that's the problem. The problem is Microsoft has to believe a legally adverse interest exists, and I don't see any basis for Microsoft to have any fear that it was going to be sued by DataTurn. I don't think the standard can or should be shaking in your boots. The question is their software is being challenged and accused in cases against all their customers that are generically being stated customers. Okay, but listen. Say Microsoft sells a screw, a screw, just a screw. The claim is to a bicycle, a wheel connected by that screw to the frame. Okay? Do I have your attention? Yes. All right. I don't know if you're curious, but I'd like to. Wheel, screw, frame. All right? And suppose that I am the patentee, and I sue Microsoft and say, when you sell people this screw, you know that they are putting it together with a wheel and a frame and making a bike that infringes my claim. Okay? Do I now have an indirect infringement claim against Microsoft for selling the screw? I need to know more facts. No, you don't, because those are the only facts you have here. The only facts you have here is data-turned claim construction charts that indicate that addo.net is being used by customers in a manner that amounts to direct infringement. That's it. That's all that's present here. Do the indemnity requests and the pattern of litigiousness add anything to this? I was going to say the second problem I was going to hope to get to is indemnity. I mean, I would say on the first that if you read a claim against every single element is satisfied by a product, and you sue many customers, that that is sufficient to mean that someone could bring in a direct infringement claim. But that would be suggesting that all you have to have is direct infringement to bring a contributory claim. That can't possibly be the standard. I think that if a supplier supplies a product, and it goes to many customers, and they all get the same claim chart, that the implication is by selling the product to all those people that they're inducing infringement, and that that product is substantially designed for the infringement. I believe those inferences are fair. I think that should be enough to get you in the courtroom. I think that meets the test. I mean, that's what I think. Especially given the litigious activity and the indemnity demand. Were they ever litigious against an indirect infringer? Because I only saw them going after people who used the product in a way that amounted to direct infringement. I didn't see them being litigious against the supplier. Right. It's very common. And there are millions of the fact patterns where they just want to sue the customers on a mass basis and never let the manufacturer come into the case. Never let. Hold on. Couldn't you just have agreed to indemnify any one of these people in Texas? This isn't a case about you getting in court. This is a case about you choosing the court you want to be in. You could have agreed to indemnify any one of those Texas customers and stepped into their shoes in Texas. Couldn't you? And then try to join the case that we sent down there? You could have totally defended it. You could have taken over the case. So that goes to the question of whether indemnity that's disputed should be enough to be an adverse legal interest. And I believe that it clearly should be. It seems to me that you don't have to – requiring someone to accept that kind of obligation broadly when there's so many disputed issues. The whole point is to eliminate risk. The question to me is, is it right? It comes to me a rightness issue. Is there enough information such that a court can just – is able to just – is it justiciable? But you realize if you agreed to indemnify or if you were required to indemnify, you could have only gone in Texas because there's already a suit there. You wouldn't have had any basis to file in New York. You'd have to accept the full indemnity risk in order to do that. So how can the sector of indemnity allow you to get into New York but not force you to litigate in Texas where the case is already ongoing? Because I don't think it's – it's not the same case. So if it's not the same – It's a broader case. Okay. So if it's not the same case – It's the real parties and interests actually litigating. Okay. So what you're saying is if we agreed to indemnify, we could then only step into the shoes of the direct infringer and argue their case, correct? So how does that give you a basis to claim indirect infringement and go up north? I think the issue is ripe and justiciable when you still can test indemnity. I don't think you have to accept indemnity before you can defend your profit. If you can prove that your product is not infringing, that should be enough without litigating indemnity everywhere. The whole point is to be able to free yourself from those kind of legal risks. That's what I think Benemune says. That's what other Supreme Court cases say. I don't think you have to accept all that indemnity. All right. Thank you, Mr. Rice. Thank you. We'll hear from you in the next case as well. Mr. Belt, you have rebuttal. Thank you. It's a matter of policy. Don't we prefer a single suit against the supplier rather than letting you sue dozens and dozens of customers and avoid the person that you're ultimately after? Each case, as I understand it, was— Every single one of these claim charts lists the Microsoft product. Who are you really after here? The focus was on the end users, the defendants in the taxes case. Yeah, but we understand that if you can sue a hundred mom-and-pop shops and threaten them with very expensive litigation, you're going to get a lot of settlements and find yourself in a lucrative patent-milking business. Why don't you bring it right to the head and sue Microsoft right up front? Because they'll probably contest you. Is that what's going on here? That is not what is going on here. How do we know that? Why didn't you bring Microsoft in if you don't want that inference to be drawn? When you look at the claim charts, Your Honor, the focus was on the products and the methods that were being used by the end users To use Judge Morris' analogy, ADO.net, for example, was a tool, like a screw. But the defendants in the Texas cases were not mom-and-pop shops. Many of them were large companies who had their own IT departments that were building their own systems. And the focus, in fact, Data Terms Representative said to Microsoft, we're not interested in suing you. Just as a matter of efficiency, why don't we have one suit against the supplier rather than hundreds of suits against the users? Because there's that saying in, forget the Russian novel, every happy family is happy in the same way. Every sad family is sad in a different way. Every one of those defendants is infringing in a different way by using their own products and their own methods and systems. So you're saying there's absolutely no basis for an indirect infringement charge here against Microsoft? No, we later discovered that there was, but at the time, you have to focus, as Judge Morris said, at the time that Microsoft filed its complaint, what were the state of the facts? Were you willfully blind at that point? Could you have taken a step to find out whether there was going to be the liability that Judge Proch refers to? We weren't willfully blind. I don't think there's any accusation of that, so I'm not sure about that. What Mr. Killa said, he was the 30B6 witness, he said, I looked at those claim charts that the defendants in Texas were sending us. Now I didn't see the Microsoft code being implicated there. I'm just talking about the source code. That was the state, and what he said was, what is being implicated there are the products that are being built by the defendants. Now, if Microsoft felt that it needed to step in, the proper step should have been to intervene, and that is I think the teaching case that we cite in our brief, this court's case from 2006. That would have been the efficient way to do it rather than, I'll say it, forum shopping to get into a different jurisdiction. But wait a minute, you're talking about the indemnification piece of this, right? Not the indirect infringement piece of this. When you're saying Microsoft should have stepped in, they should have stepped in to the Texas cases because there was an indemnification piece of this, right? It could have been either way, Your Honor. They've denied that there was an indemnification piece. Even if they felt that they were somehow, I forget who used the term, shaking in their boots. But even if they felt that they were next up, there are cases that are going on, they could have intervened in the Texas cases. They didn't. They could have intervened as an indemnification component? I mean, what would be the basis? Actually, you're right. I don't know that they could get in if they're not an interested party. And economic interest is indeed insufficient. Right. Actually, you're right, Your Honor. They could not have intervened absent the... So they would have had accepted liability in order to intervene. Indemnification. Exactly, yes. Or at least make some accommodation with their end users and customers. If I may just comment very briefly back to the normalization. One sentence. Logical table is not introduced in Column 4. It is introduced in Column 1. And the invention is about using logical tables from the... And it can start from the normalization process. I mean, I'm sorry, it can start from already normalized tables. The normalization process, as Column 1 says, allows you to define different logical tables. Not logical tables, period, but different. In other words, logical tables coming from an already normal table in the physical database. Or new ones as a result of normalization. Thank you, Your Honor. Thank you, Mr. Bell. Our next case.