[00:00:00] Speaker 04: Our next case is Sequoia Technology versus Dell et al, consolidated with Red Hat versus Sequoia et al, 2021, 22, 63, 64, 65, 66, 67. [00:00:17] Speaker 04: I trust I haven't omitted any. [00:00:20] Speaker 04: Mr. Yanku. [00:00:23] Speaker 01: Good morning, Your Honors. [00:00:24] Speaker 01: May it please the court, Andre Yanku from IRL and Manila on behalf of Sequoia Technologies [00:00:30] Speaker 01: The central issue here for claim construction is whether a logical volume of memory on a disk must contain only whole or entire disk partitions, as Red Hat argues, and as the district court found, or whether they can also contain portions of a disk partition, as Sequoia argues. [00:00:56] Speaker 01: The words and the concepts of whole or entire are nowhere to be found in the claims. [00:01:02] Speaker 01: And the district court was wrong to add them. [00:01:04] Speaker 04: You might turn down the volume a little. [00:01:06] Speaker 01: Sure. [00:01:08] Speaker 01: We got here because Red Hat. [00:01:12] Speaker 05: Could I be clear about one thing? [00:01:14] Speaker 05: Yes. [00:01:15] Speaker 05: If we were to agree that claim one is not infringed, then claim eight would not be infringed either, right? [00:01:25] Speaker 01: Well, if we're speaking about the same claim, the claim construction issues that appear in claim one regarding this partition and logical volume are the same for claim eight. [00:01:41] Speaker 01: So for that reason, the answer is yes. [00:01:45] Speaker 01: they stand or fall together on that specific point. [00:01:49] Speaker 01: Claim eight obviously has an additional issue surrounding claim section 101. [00:01:56] Speaker 01: But the main issue on claim construction, the reason we got here to what I believe is the district court's wrong construction is because Red Hat offers a false choice for constructing or configuring logical volumes [00:02:14] Speaker 01: Either, they argue, we have to construct logical volume from whole or entire partitions, or we must construct them from individual extents. [00:02:27] Speaker 01: The patent rejects the second option, so Red Hat argues we must have the first option, which is whole or entire partitions. [00:02:36] Speaker 01: But that's wrong. [00:02:39] Speaker 01: The invention here is not about either option one or option two. [00:02:44] Speaker 01: The invention here instead provides a combined approach. [00:02:49] Speaker 01: The logical volume is constructed or configured or managed at the disk partition level while storage space is allocated at the extent level. [00:03:06] Speaker 01: So here is the confusion that we must clear up. [00:03:10] Speaker 01: Construction of logical volumes is not the same as allocation for storage purposes. [00:03:20] Speaker 01: The patent makes it clear, the specification makes it clear that disk partitions are the construction units while extents [00:03:30] Speaker 01: are the space allocation units. [00:03:33] Speaker 03: So can I ask you, I hear what you're saying. [00:03:35] Speaker 03: You're saying that you construct the logical value based on the disk partitions, and then you can decide, I guess, using the extent allocation table, which extents will be in the logical value or not. [00:03:50] Speaker 03: Correct. [00:03:51] Speaker 03: It's a little bit not expressly clear from that claim. [00:03:54] Speaker 03: I mean, I would say this. [00:03:55] Speaker 03: You've got a construction for extent allocation table that you're disputing. [00:04:00] Speaker 03: And the question is whether, of course, that means when it says that extent is used or not used, whether that means it's actually being used to store or whether it's part of the logical value, right? [00:04:11] Speaker 01: That's absolutely right. [00:04:13] Speaker 01: And that is absolutely key to understand. [00:04:15] Speaker 03: What is your best argument for why I should understand that phrase to be the way you interpret it? [00:04:22] Speaker 01: Well, first of all, the words of the claim itself. [00:04:24] Speaker 01: It's an extant allocation table. [00:04:27] Speaker 01: It's not an extant storage table. [00:04:30] Speaker 01: So that's number one. [00:04:31] Speaker 01: Number two, again, look at the spec. [00:04:34] Speaker 01: Extant is, quote, the minimum unit of space allocation to store information. [00:04:41] Speaker 01: OK, it's not, as Red Hat often does in its papers, say it's not the minimum unit of space to store information. [00:04:51] Speaker 01: We can just simply write out the word. [00:04:53] Speaker 03: Is there some extrinsic evidence where the word allocation was used to mean, I realize it's extrinsic evidence, but nonetheless, there is some evidence supporting the idea that allocation is not referring to storage. [00:05:06] Speaker 03: I mean, it is referring to storage. [00:05:07] Speaker 03: that is used synonymously with storage, right? [00:05:11] Speaker 01: So what we have here is a virtualization. [00:05:15] Speaker 01: The whole scheme that's patented here is about virtualization for space on disk for storage. [00:05:26] Speaker 01: In other words, it's not the actual storage itself. [00:05:30] Speaker 01: There is an inventor paper [00:05:33] Speaker 01: that is in the record, that the district court relied on, said it's part of the intrinsic record. [00:05:38] Speaker 01: And that actually makes it clear. [00:05:41] Speaker 01: And you can look at the inventor paper. [00:05:43] Speaker 01: It's the joint appendix 796, 798, thereabouts. [00:05:48] Speaker 01: And it makes it clear that there are two different types of functions here. [00:05:56] Speaker 01: there is a file system, and figure one of the patent also makes it clear, there's a file system above the logical volume system. [00:06:06] Speaker 01: The file system actually does the storage and knows and keeps track of what is in each extant on the disk. [00:06:19] Speaker 01: When we talk about the logical volume, [00:06:23] Speaker 01: which is in the claims. [00:06:25] Speaker 01: That is the virtualization layer. [00:06:27] Speaker 01: And that's where we talk about allocation. [00:06:31] Speaker 01: and that the file system later can use or not use that specific space. [00:06:38] Speaker 03: What about the fact that the specification has one embodiment in which it really is true that in that embodiment, at least, for certain, the smallest possible unit in the logical value is the disk partition, right? [00:06:55] Speaker 03: You agree with that, right? [00:06:56] Speaker 03: There is an embodiment. [00:06:57] Speaker 01: For sure, but it's the smallest unit [00:07:02] Speaker 01: for construction of the logical volume, it is not the smallest unit for space allocation. [00:07:10] Speaker 01: At column seven, lines two and three, so the smallest unit for construction is at the bottom of column six, then at the very next paragraph, top of column seven, says that the smallest unit for space allocation [00:07:24] Speaker 01: for storage is actually the extent. [00:07:27] Speaker 01: So you combine that with the extent allocation table. [00:07:29] Speaker 01: The only purpose of all that is to tell us whether each extent is in or is out. [00:07:35] Speaker 03: Perhaps. [00:07:37] Speaker 03: Let me ask you this. [00:07:38] Speaker 03: You're saying that there's two ways in which the logical value is constructed. [00:07:42] Speaker 03: First, partitions, and then by the extent allocation table. [00:07:46] Speaker 03: That's not necessarily clear from your spec to me. [00:07:49] Speaker 03: And so do you want to tell me where you think the best support is? [00:07:52] Speaker 03: Is there some sentence that says we define the logical value by these two mechanisms? [00:07:59] Speaker 03: Because what I see mostly in this spec, and even in the claims, is that the logical value is defined [00:08:05] Speaker 03: not by the extent, but by the partitions alone? [00:08:11] Speaker 01: Well, it is constructed at the partition level. [00:08:14] Speaker 01: That's where the metadata is configured and kept. [00:08:18] Speaker 01: And then the space allocation, what's in and what's out, is at the extent level. [00:08:23] Speaker 01: Let me give perhaps an analogy here. [00:08:26] Speaker 01: I hate [00:08:27] Speaker 01: The use of analogies in general, they get us in trouble. [00:08:30] Speaker 03: You haven't told me in the spec where the language is and support what you've said, so I don't want you to forget about that, because it's important to me. [00:08:36] Speaker 01: Well, sure. [00:08:36] Speaker 01: At the bottom of column six, again, it says there are three virtualizations. [00:08:44] Speaker 01: At column six, line 55, there can be three possible virtualizations of storage. [00:08:50] Speaker 01: Then at line 57, the first virtualization is a disk partition. [00:08:56] Speaker 01: And then it continues at line 60, a disk partition is a minimum unit of the logical volume. [00:09:01] Speaker 01: The logical volume includes more than one disk partition and so forth. [00:09:06] Speaker 01: Then the second virtualization is the logical volume. [00:09:11] Speaker 01: But then at the top of page seven, it talks about the third virtualization, which is an extant. [00:09:16] Speaker 01: The extant is continuous space having the same size. [00:09:19] Speaker 01: It is also a minimum unit of space allocation to store information. [00:09:26] Speaker 01: So there are three different virtualizations. [00:09:29] Speaker 01: Dispartition, logical volume, and extent. [00:09:31] Speaker 03: And then obviously... What does that mean? [00:09:33] Speaker 03: That it's a minimum unit of space allocation to store information? [00:09:38] Speaker 03: I mean, I'm just... You know, because the whole question is whether it's for storage or whether it's for allocation. [00:09:43] Speaker 03: This sentence uses both words. [00:09:45] Speaker 01: Yeah, yeah. [00:09:45] Speaker 01: So it's provided and it's available. [00:09:49] Speaker 01: It's in the logical volume now. [00:09:51] Speaker 01: If it's in, if that extent is in the logical volume, it's available for the file system to store data there and address data to that extent. [00:10:01] Speaker 01: If it's not allocated, then it's not in the logical volume at all. [00:10:07] Speaker 01: If OK with the court, I want to give an analogy of the risk of going awry, because analogies often do. [00:10:17] Speaker 01: But the way to think about this is the way addresses, because we're talking about addressing. [00:10:21] Speaker 01: That's the main issue here in this patent. [00:10:24] Speaker 01: Addresses, the way cities organize addresses. [00:10:30] Speaker 01: The way you construct a city is by first saying what streets are in the city, and then the numbers. [00:10:35] Speaker 01: The homes are the dwelling units. [00:10:36] Speaker 01: on a given street you could construct a city without having street names and instead each home is given its own individual identity and you know you could give it each a unique identifier but that's unscalable it's unwieldy [00:11:01] Speaker 01: So every time we have a city, we say first the street that's in the city and then the actual numbers, home numbers, that are on the dwelling units that are on that street. [00:11:13] Speaker 01: You don't have to use the whole street. [00:11:15] Speaker 01: A city sometimes ends and then the next neighboring city begins with different home numbers on the same street. [00:11:22] Speaker 01: New Hampshire Avenue goes into Maryland and so on. [00:11:25] Speaker 01: The same thing here. [00:11:27] Speaker 01: You need to construct the volume, and that's the invention. [00:11:32] Speaker 03: So you're saying first the streets, then which houses on the street are included. [00:11:35] Speaker 01: Exactly. [00:11:37] Speaker 03: Let me ask you this. [00:11:38] Speaker 03: Under your construction, there's a decent argument that your interpretation would exclude the preferred embodiment from the claim, because to the extent allocation [00:11:51] Speaker 03: They wouldn't have an extent allocation table in the preferred embodiment that does something other than just identify whether something has something stored in it or not. [00:12:01] Speaker 03: Because in the preferred embodiment, it says that it's only the partitions that are the smallest logical unit or smallest unit in the logical diagram. [00:12:12] Speaker 03: So how do you respond to that? [00:12:14] Speaker 01: Again, the extant is the smallest unit of storage. [00:12:18] Speaker 01: It's not of storage allocation. [00:12:20] Speaker 01: It's not the partition. [00:12:22] Speaker 03: So are you saying it's your view that the claim would not, under construction, the claim would not exclude the preferred embodying? [00:12:29] Speaker 03: Absolutely not. [00:12:30] Speaker 03: Why not? [00:12:31] Speaker 01: Well, again, the preferred embodiment is we first identify which partitions are in. [00:12:39] Speaker 01: And then after that, which extents within that partition are in. [00:12:43] Speaker 01: So the smallest unit that constructs the volume, we have to say it's partition one, partition two, and so on. [00:12:51] Speaker 01: And then within that, which extents. [00:12:53] Speaker 01: That's in the specification. [00:12:55] Speaker 01: And that also is what the claim requires under our construction. [00:12:59] Speaker 03: flying the face of the language and the specification that says, in the preferred environment, the partition is the smallest unit by which the logical volume can be formed. [00:13:15] Speaker 01: The smallest unit. [00:13:16] Speaker 01: Say that again. [00:13:16] Speaker 01: I'm sorry. [00:13:17] Speaker 03: I'm sorry. [00:13:17] Speaker 03: I thought that in the preferred environment, the partition, or the extent, was the not the extent. [00:13:26] Speaker 03: I'm sorry. [00:13:27] Speaker 03: Let me try that again. [00:13:30] Speaker 03: I might ask you this question. [00:13:32] Speaker 03: Oh, that the disk partition is the smallest [00:13:36] Speaker 03: unit of the logical value, which means that you could not then separate it further by identifying some extents that are in and some extents that are not. [00:13:47] Speaker 01: It's the smallest unit of construction. [00:13:49] Speaker 01: It's bottom of column 11. [00:13:51] Speaker 01: This partition is the volume construction unit. [00:13:55] Speaker 01: But again, at the top of column 7, the extent is the minimum unit of space allocation. [00:14:01] Speaker 01: Let's assume that each partition has four extents. [00:14:04] Speaker 01: Each extent is one megabyte. [00:14:06] Speaker 01: So each partition has four megabytes. [00:14:08] Speaker 01: And at configuration, the request is to have a 10 megabyte logical volume. [00:14:14] Speaker 01: What's the technical reason to not allow that? [00:14:18] Speaker 01: So if we have 10 megabytes, you have to have two full partitions, four plus four. [00:14:25] Speaker 01: And then the third partition [00:14:26] Speaker 01: you have extent one and extent two. [00:14:29] Speaker 01: And then three and four are not allocated to this logical volume. [00:14:36] Speaker 03: I want to ask you about that. [00:14:40] Speaker 03: Are you aware of any cases in which the plain and ordinary meaning of a claim term is determined by looking at other patent specifications? [00:14:50] Speaker 03: I mean, I'm specifically talking about the CRM issue now, just to give you context. [00:14:55] Speaker 03: But there, there is reliance on many 36, I think, other patents. [00:15:00] Speaker 03: that in those cases may have used lexicography to define CRM a particular way. [00:15:05] Speaker 03: And that's being used now here in this case to say what the plainly ordinary meaning of that word is. [00:15:10] Speaker 03: Are you aware of any cases where we have not such an approach? [00:15:14] Speaker 01: I am not. [00:15:15] Speaker 01: I am not. [00:15:15] Speaker 01: And in fact, that shows the exact opposite. [00:15:17] Speaker 01: The fact that those older patents [00:15:20] Speaker 01: needed to be their own lexicographers to define CRM as to include in their specification, to include signals, transitory signals, just shows that that's not the plain and ordinary meaning. [00:15:37] Speaker 01: Here, what's clear is that the patent itself says absolutely nothing about signals. [00:15:45] Speaker 01: The context of the invention is about storage on disk. [00:15:49] Speaker 01: And it's not transitory. [00:15:51] Speaker 01: It has to be permanent memory. [00:15:54] Speaker 01: And the key to this issue is the controlling authorities' mentor graphics. [00:16:03] Speaker 04: Judge Stein has a question. [00:16:04] Speaker 05: Yes. [00:16:05] Speaker 05: I'm confused about the stipulation of non-infringement. [00:16:08] Speaker 05: If you look at A44, it says the accused products can form logical volumes utilizing units smaller than the whole entire disk partitions. [00:16:19] Speaker 05: But can the accused products form logical volumes using partitions, or can it do both? [00:16:27] Speaker 01: Yes, again, so the accused products form [00:16:33] Speaker 01: disk form logical volumes by first identifying the disk partition, and then they allocate space to that logical volume from each disk partition on an extent by extent basis. [00:16:48] Speaker 05: So they're always using extents to form the logical volume. [00:16:51] Speaker 01: You're forming the logical volume, you're constructing it by naming the disk partition, naming the street, and then [00:17:02] Speaker 01: You allocate space to that memory by identifying which extent is in and out. [00:17:10] Speaker 01: It's a two layer. [00:17:11] Speaker 05: And that renders it non-infringing? [00:17:15] Speaker 01: No, no. [00:17:16] Speaker 01: That would infringe under our construction. [00:17:18] Speaker 05: No, no, under the construction that was adopted here by the district court. [00:17:25] Speaker 01: Right. [00:17:25] Speaker 01: Under the court construction, it doesn't infringe, because under the district court's construction, there is a gloss that's given by the district court that you have to use, you have to allocate all the extents in a given partition. [00:17:42] Speaker 01: In my city example, you would have to construct a city where every street [00:17:47] Speaker 01: Every home on every street is always in that city, and that street cannot continue into the next city. [00:17:54] Speaker 01: So that's the court's construction. [00:17:58] Speaker 01: Mandating, it's not in the claims. [00:18:00] Speaker 01: The court's construction mandates that every bit of memory in every dispartition is allocated [00:18:10] Speaker 01: to that logical volume. [00:18:11] Speaker 01: It's a concept that's simply not in the claims. [00:18:13] Speaker 03: Just to make sure we understand the stipulation. [00:18:15] Speaker 03: The reason why there's a stipulation of non-infringement is because the accused products don't do that. [00:18:19] Speaker 01: Correct. [00:18:20] Speaker 01: Because the accused products don't use the entire disk partition every single time. [00:18:27] Speaker 01: They use sometimes. [00:18:27] Speaker 03: They can pick half partitions or half the houses. [00:18:32] Speaker 00: Yeah. [00:18:32] Speaker 00: Correct. [00:18:33] Speaker 00: Which, again, we believe is what the claims allow as well. [00:18:40] Speaker 04: You have, of course, consumed our time, but we'll give you three minutes for a bottle. [00:18:44] Speaker 01: Thank you very much, Your Honor. [00:18:45] Speaker 04: Mr. O'Quinn. [00:18:47] Speaker 04: Thank you. [00:18:48] Speaker 04: If you need a little more time, we'll provide it. [00:18:50] Speaker 02: Thank you, Judge Blory. [00:18:51] Speaker 02: I may please the court, John O'Quinn on behalf of the appellees. [00:18:55] Speaker 02: Judge Stike, with respect to the question that you asked just a moment ago, the reason there's a stipulation of non-infringement is that the accused products do not use disk partitions as a volume construction unit. [00:19:07] Speaker 02: They don't use disk partitions as the minimum unit of the logical volume. [00:19:12] Speaker 02: Judge Stole, you had asked a question earlier about [00:19:14] Speaker 02: was their language in the specification that talked about using the disk partition as the minimum unit of the logical volume. [00:19:22] Speaker 02: And there is. [00:19:23] Speaker 02: That's a column 6, line 60 to 63. [00:19:26] Speaker 02: And the court says, therefore, a size of the logical volume is resized in disk partition units. [00:19:33] Speaker 02: And then similarly, at column 11, starting at line 66, [00:19:38] Speaker 02: the present invention, which of course is definitional language, constructs a logical volume by using a disk partition as a volume construction unit so that the present invention can minimize the size of metadata. [00:19:51] Speaker 02: And so this, of course, the principal issue is whether this specific claimed method of forming a logical volume in contrast to methods that were in the prior art [00:20:01] Speaker 02: which the patent owner sought to distinguish in front of the patent office, whether this method involves gathering disk partitions and forming logical volumes out of disk partitions as the claims, as the specifications say, as they distinguished from the prior art, or something less than partitions, like extents. [00:20:22] Speaker 03: I want to ask you a question. [00:20:24] Speaker 03: The quote that you had in column 11 going on to column 12, [00:20:27] Speaker 03: That's the only place I see in the spec that talks about how you achieve the goal of reducing the size of metadata. [00:20:34] Speaker 03: Are you aware of anything else in the specification that talks about how you achieve that goal? [00:20:39] Speaker 02: So that's my understanding of where it links the goal of minimizing metadata to using disk partitions as a volume construction unit. [00:20:47] Speaker 02: There are other places in the specification that talk about the disk partition [00:20:52] Speaker 02: being the minimum unit, but I'm not aware of other places in the specification that specifically make the link. [00:20:58] Speaker 03: There could be some other reason why metadata is reduced, but this is the only reason I see. [00:21:02] Speaker 03: Do you think there's any other reasons in the specification? [00:21:05] Speaker 02: I don't think there's anything else in the specification that points to a different way in which the claimed invention purports to reduce metadata other than by using disk partitions. [00:21:15] Speaker 02: Because as we illustrated at page 13 of our brief, if disk partitions are being shared [00:21:22] Speaker 02: across multiple logical volumes, then that means that you're necessarily going to balloon the metadata. [00:21:27] Speaker 02: Because they're going to have to have metadata not just for, that is, the disk partition is going to have to have metadata not just for one logical volume, but for multiple logical volumes. [00:21:38] Speaker 03: Does the use of an allocation table somehow reduce the amount of metadata? [00:21:42] Speaker 02: Well, the use of the allocation table is really serving a different function, Judge Stoll. [00:21:49] Speaker 02: I think that the extent allocation table is telling whether or not the extent has been allocated for the storage of information. [00:21:58] Speaker 02: There are really only two places [00:21:59] Speaker 02: in which the extent is described. [00:22:01] Speaker 02: Tellingly, extents aren't even mentioned anywhere in the summary of the invention. [00:22:05] Speaker 02: The two places in which extents are described in the patent are column 7, lines 1 through 5, and column 8, line 36. [00:22:13] Speaker 02: Both of them describe the extent as being the minimum unit of space allocation to store information. [00:22:20] Speaker 02: And so the extent allocation table [00:22:22] Speaker 02: is telling you that that extent has been allocated by the logical volume, not to the logical volume, but by the logical volume for the storage of information. [00:22:34] Speaker 02: That is, information is actually being stored in it. [00:22:38] Speaker 02: And the district court, this was Judge Stark's opinion at Appendix 36, said, quote, the patent specification describes an extent as the minimum unit of space allocation to store information [00:22:49] Speaker 02: thereby directly linking an extents purpose to storing information. [00:22:56] Speaker 02: And similarly, there is no example within the patent of an extents being used the way that they describe. [00:23:03] Speaker 02: That is, of gathering extents, which it would have been very easy to say, and using extents to construct [00:23:10] Speaker 02: a logical volume. [00:23:14] Speaker 03: What about the three virtualizations, including the language at the top of column seven, lines one through five? [00:23:22] Speaker 03: The third virtualization is an extent. [00:23:27] Speaker 03: How do you respond to that? [00:23:28] Speaker 03: That's not a bad quote there. [00:23:30] Speaker 02: Well, I think what's going on is this is simply defining what an extent is, [00:23:35] Speaker 03: What about virtualization? [00:23:38] Speaker 03: We kind of understand what virtualization means. [00:23:43] Speaker 02: What it's saying by virtualizations here is the logical volume is a virtualization. [00:23:47] Speaker 02: The file manager, the file system sees that as though it was a hard drive. [00:23:51] Speaker 02: Disk partitions themselves are actually a virtualization. [00:23:54] Speaker 02: An operating system goes in and overlays the partition on top of the actual physical disk. [00:24:01] Speaker 02: The extent is also a virtualization. [00:24:04] Speaker 02: It doesn't exist in nature. [00:24:06] Speaker 02: It's when the disk partitions are gathered, what column 7 at the top tells you is the size of the extent, like how big your zip code is going to be, is decided by the logical volume. [00:24:19] Speaker 02: It's decided at the creation of the logical volume. [00:24:22] Speaker 02: Extents can vary in size. [00:24:24] Speaker 02: They're not a standard unit. [00:24:25] Speaker 02: You can have one logical volume that has an extent that's one size and another logical volume that has extents that are other size. [00:24:31] Speaker 02: And so at the risk of taking a slightly different analogy, the United States is constructed out of states. [00:24:38] Speaker 02: States are like disk partitions. [00:24:41] Speaker 02: Overlaid on top of that are zip codes. [00:24:43] Speaker 02: But the United States isn't constructed out of zip codes. [00:24:45] Speaker 02: That is just a way of organizing information. [00:24:48] Speaker 02: Well, the same thing is true here. [00:24:49] Speaker 02: The logical volume manager gathers disk partitions. [00:24:55] Speaker 02: That is the construction unit, and it gathers at the disk partition level. [00:25:00] Speaker 02: But then, extents are overlaid as a way for determining what amount of space an extent is going to have, so that when a file needs to be written to a particular logical address. [00:25:10] Speaker 05: You mean a partition that's going to happen. [00:25:12] Speaker 05: I'm sorry, just saying. [00:25:12] Speaker 05: You mean a partition that's going to happen. [00:25:14] Speaker 02: A partition is gathered, the disk partitions are gathered, and then the extent, the size of the extents are overlaid on top of that. [00:25:23] Speaker 05: I think you misspoke. [00:25:24] Speaker 05: You meant the analysis of the extent shows how much space is available in the partition. [00:25:30] Speaker 02: Yes. [00:25:30] Speaker 02: Well, in logical volume, but also in the partition. [00:25:33] Speaker 02: Because as Sequoia specifically argued to the Patent Office in order to avoid institution, [00:25:44] Speaker 02: While the logical volume is formed from extents, I'm quoting in appendix 921, extents are added or removed from a logical volume at the level of disk partitions. [00:25:56] Speaker 02: And if you read the next sentence, it explains that even further. [00:25:59] Speaker 02: It says, that is, disk partitions are gathered to form a logical volume, and thus, thus, the extents, excuse me, the extents of the gathered disk partitions form the logical volume. [00:26:11] Speaker 02: In other words, [00:26:12] Speaker 02: You take the disk partition, and all of its extents are what are used in the logical volume. [00:26:21] Speaker 02: And the patent office understood that that was the patent owner's argument, because in construing the claims before it, the patent office, this is at appendix 967, made clear that says that with the understanding that [00:26:36] Speaker 02: The language form a logical volume refers to actual storage space allocated exclusively to a logical volume. [00:26:44] Speaker 02: That is that in the claim, all of the extents of a disk partition are allocated exclusively to one logical volume. [00:26:54] Speaker 02: And they went on to say, quote, the specification of the 436 patent [00:26:58] Speaker 02: does not include any overlap in the disk partitions of separate logical volumes." [00:27:04] Speaker 02: End quote. [00:27:04] Speaker 02: Again, that's Appendix 967. [00:27:06] Speaker 02: So the patent office understood that the patent owner's argument was that extents [00:27:13] Speaker 02: are allocated at the disk partition level, and all of the extents associated with the disk partition can be allocated to one and only one logical volume. [00:27:24] Speaker 03: One thing you said that was interesting and gathered my attention was you said the logical volume is the thing that's allocating the extents. [00:27:30] Speaker 03: Where is your support with that? [00:27:33] Speaker 03: Is there a particular sentence that phrases it that way, somewhere in the intrinsic evidence? [00:27:38] Speaker 02: Well, I think first I'd point to column 7, line 1 through 5, and column 8, line 36, both of which talk about the extent being the minimum space allocation unit. [00:27:50] Speaker 02: And that is that it is saying this is the minimum. [00:27:53] Speaker 03: I understand. [00:27:53] Speaker 03: But I wanted to know if you had something specifically about the logical value being the one that's performing the allocation. [00:28:00] Speaker 02: I don't have anything more specific than this in the same way that [00:28:05] Speaker 02: I mean the thing that I do have of course is that if you look at column 11, lines 50 to 51, the specification tells you how the size of [00:28:18] Speaker 02: Logical volumes are changed dynamically, and that is by, quote, adding new disk partitions. [00:28:24] Speaker 02: And there's not a single example in the specification of saying that the size can be changed by adding or removing extents. [00:28:32] Speaker 02: And the only thing they cite. [00:28:34] Speaker 03: The claim language, though, is very clear. [00:28:37] Speaker 03: Let's just say the claim language is very clear. [00:28:38] Speaker 03: And it said exactly what you just said. [00:28:40] Speaker 03: The logical volume can be changed by adding or removing extents. [00:28:45] Speaker 03: In that case, you know. [00:28:47] Speaker 03: Don't you think that that could be claimed, even though it's maybe not a preferred embodiment? [00:28:52] Speaker 03: And we're getting into written description now, but what you just said just didn't make sense to me. [00:28:57] Speaker 02: Oh, my only point, Judge Stoll, was that there's no example in deciding how to construe the claim language. [00:29:04] Speaker 02: There's no example in the specification of constructing use at the extent level as opposed to constructing and resizing at the disk partition level. [00:29:13] Speaker 03: So you're saying that supports your understanding of it? [00:29:15] Speaker 02: Correct. [00:29:15] Speaker 02: The other thing that supports my understanding of the extent allocation table tracking usage is that is then how the logical volume can know how to resize. [00:29:28] Speaker 02: Because resizing isn't just increasing by adding [00:29:31] Speaker 02: of material to it, it is also decreasing. [00:29:34] Speaker 02: If you look at column four, line 48, it refers to decreasing the size of a logical volume. [00:29:40] Speaker 02: Well, if you're not tracking, if the logical volume manager is not tracking which of its extents are in use, that is to say are actually storing information, [00:29:50] Speaker 02: then it's going to be potentially giving up physical space that is actually what is holding the data that the file system ultimately wants to access. [00:29:58] Speaker 02: Because the file system has no idea where it is on the physical space. [00:30:02] Speaker 02: The file system [00:30:03] Speaker 02: just knows its logical address. [00:30:06] Speaker 03: Do I remember correctly that one of the arguments being made by Sequoia, I think, is that the file system identifies whether a particular extent is being used for storage or not? [00:30:20] Speaker 03: And it's not the allocation table. [00:30:22] Speaker 03: How do you respond to that? [00:30:23] Speaker 02: So I respond to that in two ways, Judge Stoll. [00:30:25] Speaker 02: First of all, this unclaimed file manager seems to be getting a lot of primacy in their arguments. [00:30:32] Speaker 02: And none of what they are attributing to this unclaimed file manager is anywhere to be found in the specification. [00:30:39] Speaker 02: Second, to the point that I was just making, I think if you look, for example, at column 6, line 28, what you see is that the file system manager [00:30:48] Speaker 02: it's acting on the logical volume. [00:30:51] Speaker 02: But I think all parties agree that the file system manager doesn't know the physical location of the data that is ultimately being written down at the physical level. [00:31:02] Speaker 02: The only thing that knows what the physical address is, is the logical volume manager. [00:31:10] Speaker 02: So if the logical volume manager is not tracking whether or not a particular extent [00:31:17] Speaker 02: is storing information at the physical level, then there's no way for the logical volume manager to know, can I give up this disk partition? [00:31:28] Speaker 02: Can I give up this extent? [00:31:30] Speaker 02: Do I need to move this data as part of the resizing? [00:31:33] Speaker 02: And we walk through in our brief how the specification describes the logical volume manager being the one that moves information to different physical locations in order to accommodate resizing requests. [00:31:46] Speaker 05: OK, before we run out of time, I want to talk about claim eight. [00:31:50] Speaker 05: Yes, Judge Day. [00:31:51] Speaker 05: If you were to prevail on non-infringement, do you care about whether claim eight is invalid? [00:31:59] Speaker 02: Well, obviously, if we prevail on non-infringement, that covers all the claims. [00:32:05] Speaker 02: We'd obviously prefer a decision by the court on the issue that the parties have [00:32:12] Speaker 02: So that was where I was going to go with this, Judge Stoll. [00:32:17] Speaker 02: I don't know that you have to decide this under cardinal chemical. [00:32:21] Speaker 02: That would essentially be the question. [00:32:22] Speaker 05: I think we have to decide it. [00:32:23] Speaker 05: If you want us to decide it, if you don't want us to decide it, we don't have to decide it. [00:32:28] Speaker 02: Well, I put it this way, Judge Dyke. [00:32:31] Speaker 02: I mean, I don't think the cardinal chemical compels that you decide it. [00:32:36] Speaker 02: No, no. [00:32:37] Speaker 05: What do you want? [00:32:37] Speaker 05: Do you want us to decide it, or don't you care if you win on infringement? [00:32:41] Speaker 05: Well, I think we do care. [00:32:43] Speaker 05: OK, so it strikes me that the construction of plane A is questionable, because it's talking about storage. [00:32:53] Speaker 05: And I have a difficult time seeing how that can extend so far as to encompass radio waves. [00:33:05] Speaker 05: Radio waves don't store things. [00:33:07] Speaker 02: So Judge Dyke, first of all, all of these so-called Beauregard-type claims are ultimately about storing a piece of software, right? [00:33:21] Speaker 02: I mean, at the end of the day, the Beauregard recognized an exception to be printed. [00:33:27] Speaker 05: I don't think you're addressing my question. [00:33:29] Speaker 05: How can it make sense to say that it encompasses radio waves when radio waves don't store things? [00:33:37] Speaker 02: Well, first of all, Judge Dyke, this doesn't just encompass radio waves. [00:33:43] Speaker 02: It also encompasses things that are coming over wires on the internet that are transitory. [00:33:49] Speaker 02: But due for a short period of time, that physical thing does have the data, even though it is transitory. [00:33:54] Speaker 03: But going back to Judge Dyke's question, how can this invention make sense in light of something that's transitory? [00:33:59] Speaker 02: Well, so Judge Stoll, I think all of these types of claims, these computer-readable medium, computer-readable recording medium with instruction, every single- Mis-specification. [00:34:10] Speaker 03: But I don't think- Mis-specification. [00:34:13] Speaker 03: It has to be understood in light of this specification. [00:34:16] Speaker 02: So I agree with you, Judge Stoll, two issues. [00:34:19] Speaker 02: One is, to the extent that you are pointing to the language at column 11, [00:34:26] Speaker 02: lines 36 to 39. [00:34:27] Speaker 02: It's just boilerplate language that the method of the present invention can be stored in computer readable medium, including compact disc medium, random access memory, floppy disk, hard disk, magneto-optical disk. [00:34:40] Speaker 02: And there are lots of examples. [00:34:42] Speaker 02: We pointed to 34 of them to answer your question. [00:34:45] Speaker 03: But again, I don't know of any case that looks at lexicography and other patent specifications and assigns that the plain and ordinary meaning of a term. [00:34:54] Speaker 02: So Judge Stoll, I agree. [00:34:56] Speaker 02: I'm not aware of any Federal Circuit case that decided this. [00:34:59] Speaker 02: This was exactly the mode of the analysis by the PTAB in the ex parte Mayweather case that is cited in our briefing. [00:35:08] Speaker 02: And that is exactly what the Patent Office did in a presidential decision a decade ago that is still controlling, with respect to the language, machine-readable storage medium. [00:35:21] Speaker 02: And the Patent Office found just consi- Maybe they're wrong. [00:35:24] Speaker 02: Well, they could be, Judge Dyke. [00:35:26] Speaker 02: They could potentially be. [00:35:31] Speaker 02: But what I would submit is that what the Patent Office did there is what the district court did here, which is to look at extrinsic evidence in a situation where certainly the plain language could potentially encompass transitory media. [00:35:49] Speaker 02: I mean, that's been the position of the Patent Office for over a decade. [00:35:52] Speaker 02: going back to the 2010 memorandum from the patent office. [00:35:56] Speaker 03: That's before this patent was written, right? [00:35:58] Speaker 03: By the way, that memorandum just says, can be. [00:36:01] Speaker 03: I don't think it defines the term. [00:36:04] Speaker 03: It says may. [00:36:05] Speaker 02: So Judge Stone, my modest point on this is just simply that if the language can reasonably be read either way, then [00:36:15] Speaker 02: Under Teva v. Sandoz, the district court, sitting as fact finder, gets to consider the extrinsic evidence and decide, are these 34 instances of lexicography, or does this represent what a person of ordinary skill in the art would have understood at the time? [00:36:35] Speaker 05: A person of ordinary skill in the art would understand storage to mean something other than transitory storage, right? [00:36:43] Speaker 02: So I don't think that that is necessarily true, Judge Dyke, because I think that the word storage isn't doing any more work than the word medium. [00:36:51] Speaker 02: I mean, you would understand medium as being to write it down somewhere. [00:36:55] Speaker 02: Storage is doing exactly the same thing. [00:36:57] Speaker 02: That is exactly what the patent office concluded in the Mayweather case that I made reference to. [00:37:03] Speaker 02: And under Teva v. Sandoz, it is ultimately for the fact finder to weigh the extrinsic evidence and decide, is this lexicography? [00:37:11] Speaker 02: Is this something else? [00:37:12] Speaker 02: Now, Judge Stoll, you asked about this court's decision in Mentor Graphics. [00:37:17] Speaker 02: Mentor Graphics, I think, looked at and just said, well, this is easy. [00:37:20] Speaker 02: There's no question that the patent owner acted as their own lexicographer. [00:37:24] Speaker 02: And indeed, the district court in the Acceleration Bay case that's cited in the briefing [00:37:28] Speaker 02: When it reached the same conclusion that the district court did in this case, the Acceleration Bay court said, I'm not going to rely on Mentor Graphics, because Mentor Graphics decided that case based on lexicography. [00:37:40] Speaker 04: But the fact that you have- Counsel, has this patent expired? [00:37:44] Speaker 02: I don't believe that it has, Judge Leroy. [00:37:47] Speaker 04: I don't recall the dates. [00:37:49] Speaker 04: The starting date was December 2001. [00:37:52] Speaker 04: It was extended by 149 days. [00:37:54] Speaker 02: So perhaps it has, Judge Lurie. [00:37:57] Speaker 04: Is any other litigation pending in relation to it? [00:38:01] Speaker 02: I'm not aware of any, but I would defer to my colleague on the other side on that question. [00:38:05] Speaker 02: And Judge Dyke, to come back, as I said, while certainly we would welcome a decision on this, whether we want it or not, I think the court has discretion under Cardinal Chemical to decide whether or not it has to reach and decide [00:38:22] Speaker 02: the particular issue. [00:38:24] Speaker 02: Cardinal chemical involved a situation where there was litigated invalidity and a lot of different policy considerations. [00:38:31] Speaker 02: The court in Cardinal Chemical said, quote, we acknowledge that factors in an unusual case might justify the court's refusal to reach the merits of a validity determination, a determination which it might therefore be appropriate to vacate. [00:38:44] Speaker 02: Well, here you don't even have the litigated merits of a validity [00:38:47] Speaker 02: determination. [00:38:48] Speaker 04: So I think the court has discretion. [00:38:51] Speaker 04: Thank you, Judge Laurie. [00:38:59] Speaker 01: The patent has expired. [00:39:00] Speaker 04: Is there other litigation pending? [00:39:03] Speaker 01: Not right now, other than this one. [00:39:06] Speaker 01: So returning back to the claim construction point about disk partitions and logical volumes, three key points. [00:39:15] Speaker 01: Number one, there were a lot of questions about the preferred embodiment. [00:39:20] Speaker 01: There is nothing in the specification to limit the preferred embodiment or any other embodiment to using whole or partial partitions. [00:39:30] Speaker 01: It just doesn't specify it. [00:39:32] Speaker 01: one way or the other in the specification. [00:39:36] Speaker 01: But what we do know is that there is an extant allocation table and that extant allocation table, the whole purpose of it, otherwise we wouldn't have it, the whole purpose of the extant allocation table is to tell us whether extants are allocated or not allocated. [00:39:54] Speaker 01: That's the whole point of it. [00:39:56] Speaker 01: In terms of the actual storage, [00:39:58] Speaker 01: Again, I commend the court to the inventor paper, which is in the record, Joint Appendix 794 through 799. [00:40:07] Speaker 01: At page 796, it speaks very specifically about who keeps track of the actual storage, what is in each one of these extents. [00:40:20] Speaker 01: At the top of the page, [00:40:22] Speaker 01: 796, it tells us that it is in the file system. [00:40:27] Speaker 01: And the file system tells us whether the extant has data, whether the extant has no data, and what type of data, metadata or regular data. [00:40:38] Speaker 01: That's in the file system that leaves the extant allocation table in the logical volume. [00:40:45] Speaker 01: And if we look at figure one of the patent, we see that the logical volume is a step below the file system. [00:40:51] Speaker 01: The extent allocation table in the logical volume has to do something. [00:40:57] Speaker 01: And what it does is what the name says. [00:41:00] Speaker 01: It knows whether it allocates. [00:41:03] Speaker 01: And in this preferred embodiment, that's done with one bit. [00:41:07] Speaker 01: That's it, one bit. [00:41:09] Speaker 01: Zero or one. [00:41:11] Speaker 01: Is it in the logical volume or is it not in the logical volume? [00:41:15] Speaker 01: That is in the specification. [00:41:16] Speaker 01: That one bit example is in the preferred [00:41:21] Speaker 01: embodiment. [00:41:22] Speaker 01: And if we construe the claims the way the district court has construed them, we completely eliminate that embodiment and we completely eliminate the claim term and the purpose of the logical volume. [00:41:37] Speaker 01: The last point I would make is that the claim does talk about, and the patent talks about, reducing metadata. [00:41:45] Speaker 01: And there were questions about that from the court. [00:41:49] Speaker 01: There are multiple ways that metadata is reduced, but it's for any given configuration is that we have to keep in mind the metadata is reduced. [00:41:59] Speaker 01: And the main way we do that is through the mapping system that is claimed. [00:42:06] Speaker 01: At two levels, you first designate, construe with the disk partition, [00:42:15] Speaker 01: And then you identify the extents through the extents allocation table. [00:42:20] Speaker 01: There is no need to keep track of the unallocated extents. [00:42:25] Speaker 01: That reduces the metadata. [00:42:27] Speaker 01: And the translation of addresses from virtual to physical, where the continuous addresses are in the virtual world, the extents addresses are in the virtual world, and that's the [00:42:42] Speaker 01: in the mapping table, the last element of claim one. [00:42:46] Speaker 01: That is all of that together is what reduces metadata. [00:42:51] Speaker 01: I want to ask you one question. [00:42:53] Speaker 01: Yes. [00:42:53] Speaker 03: I hear what you're saying. [00:42:55] Speaker 03: The one place I saw in the patent that talks about reduction of metadata is at the bottom of column 11, going on to column 12. [00:43:03] Speaker 03: That's the language that talks about constructing a logical volume by using a disk partition as a volume construction unit so the present invention can minimize the size of metadata. [00:43:12] Speaker 03: Is there anywhere else in the patent [00:43:14] Speaker 03: that says something like, it's the mapping. [00:43:17] Speaker 03: It's both the extent allocation table and the partition that uses this side of the data. [00:43:23] Speaker 03: Do you have something to support that? [00:43:26] Speaker 01: Well, first of all, I would say it's the entire specification. [00:43:29] Speaker 01: But look at the claim. [00:43:31] Speaker 01: It's a method of managing the logical volume. [00:43:33] Speaker 03: I'm wondering in this very, very narrow question. [00:43:35] Speaker 03: I understand what you're saying. [00:43:37] Speaker 03: But in the specification, is there somewhere else other than this sentence that I've read to you that talks about how the invention minimizes the size of metadata? [00:43:48] Speaker 01: It is that sentence in the spec, your honor, but that's representative of the whole invention. [00:43:53] Speaker 01: Again, in order to be able to manage addressing, you're doing it in the metadata for the addressing. [00:44:02] Speaker 01: You're doing it, you're putting everything [00:44:04] Speaker 01: in the dispartition, at the dispartition level. [00:44:08] Speaker 01: You have to keep track of addresses using streets. [00:44:11] Speaker 01: You cannot have unique identifiers for each extent, because if you do that, city addressing becomes unwieldy. [00:44:20] Speaker 01: That's the whole point here, doing everything at the dispartition level. [00:44:25] Speaker 01: Thank you, counsel. [00:44:25] Speaker 04: I thank you for your argument, and thank both counsel cases submitted.