[00:00:00] Speaker 04: This morning I proposed to focus on Buxton's alpha blending equation, which is, we think, a clear illustration of the board's basic error in trying to split the baby based on the use of logical operators. [00:00:13] Speaker 02: Do all programs use logical operators? [00:00:16] Speaker 04: When it comes to manipulating data in a digital computer, as far as we're aware, and as far as the record shows, logical operators are always used for that. [00:00:24] Speaker 04: They're the basic building blocks of digital computing, at least for data manipulation. [00:00:30] Speaker 04: And which certainly underscores that it's going in when we don't expect these logical operators to be a patenable distinction. [00:00:38] Speaker 04: And the reason I was going to say that Buxton's alpha blending illustrates that nicely is that the two key points about alpha blending are undisputed. [00:00:48] Speaker 04: First, Buxton sets forth that arithmetic. [00:00:50] Speaker 01: Could you speak up just a little bit? [00:00:52] Speaker 04: I'm sorry. [00:00:53] Speaker 04: Thank you. [00:00:53] Speaker 04: Of course, Your Honor. [00:00:55] Speaker 04: First, Buxton's alpha blending equation [00:00:57] Speaker 04: It discloses a mathematical equation for purposes of the relevant blending. [00:01:02] Speaker 04: And second, arithmetic equations are made up of, arithmetic operators are made up of logical operators. [00:01:08] Speaker 04: Therefore, to a person skilled in the art, disclosing the arithmetic equation discloses logical operators. [00:01:14] Speaker 04: And the board determined only that there is a difference between arithmetic operators and logical operators. [00:01:21] Speaker 04: But that's true and undisputed, but also not the relevant question. [00:01:24] Speaker 04: It doesn't get you to the conclusion. [00:01:26] Speaker 04: Because the difference is only the logical operators are the basic building blocks that can be combined in various ways. [00:01:32] Speaker 04: One way they can be combined is arithmetic operators. [00:01:35] Speaker 04: So it's true that logical operators are not necessarily arithmetic, but it's not true in the opposite direction. [00:01:40] Speaker 04: Arithmetic operators necessarily entail the use of logical operators. [00:01:45] Speaker 03: I assume you're a logical operator. [00:01:48] Speaker 04: In a totally different sense. [00:01:51] Speaker 04: Obviously we're talking about how digital computers work, not how people's brains work. [00:01:55] Speaker 01: Is it your position that the arithmetic operators necessarily use logic operators? [00:02:03] Speaker 01: Correct. [00:02:03] Speaker 01: So it's an inherency argument? [00:02:05] Speaker 04: It's not inherency because the question is how the expressed disclosure would be understood by a person of skill in the art, and a person of skill in the art would understand that arithmetic operators are logical in the sense they're comprised of logical operators, and therefore that the expressed disclosure of an arithmetic operator discloses the use of logical operators. [00:02:25] Speaker 02: Is it always Boolean logic? [00:02:29] Speaker 04: No. [00:02:30] Speaker 04: The board, in fact, construed logic. [00:02:33] Speaker 04: IV had proposed the logical operator would be construed as Boolean logical operators. [00:02:37] Speaker 04: And the board rejected that, which is not on appeal. [00:02:40] Speaker 04: Instead, only the logical operator is an operator that manipulates binary data at the bit level, which obviously is as broad as can be as part of our argument. [00:02:50] Speaker 04: Yes, it is. [00:02:50] Speaker 04: It's part of our argument. [00:02:51] Speaker 04: But as the board pointed out, the logical operators may or may not be Boolean. [00:02:56] Speaker 01: uh... which is correct and but so why it's not inherently just the question is how personal skill in the art would understand the express disclosure what happens where that's disputed if you know uh... the pose and sign intellectual ventures council I presume is going to get up and say no that's not true arithmetic operators don't include logical operators then where does that leave us on a substantial evidence standard of review where there might be evidence supporting both sides [00:03:25] Speaker 04: Right. [00:03:26] Speaker 04: Well, obviously, typically speaking, if there's really evidence in support of both sides, that would normally mean there's substantial evidence that you affirm. [00:03:33] Speaker 04: The important thing here, though, is that all of the evidence is entirely consistent. [00:03:37] Speaker 04: Because they have, evidentiary-wise, they have basically three things. [00:03:41] Speaker 04: First, their expert's conclusion, which doesn't count because you read his declaration, it's just conclusory. [00:03:47] Speaker 04: It's also based on their own claim construction, which the board rejected, the limitation of Boolean. [00:03:52] Speaker 04: You then look to their two pieces. [00:03:53] Speaker 04: They have the dictionary definition. [00:03:55] Speaker 04: saying that a logical operator is non-arithmetic. [00:04:01] Speaker 04: The thing is, there are two things about that. [00:04:03] Speaker 04: First, that's not the claim construction in this case. [00:04:05] Speaker 04: The claim construction in this case is just manipulating binary data at the bit level, which arithmetic operators unquestionably do. [00:04:12] Speaker 04: I mean, if arithmetic operators do anything, they manipulate binary data at the bit level. [00:04:16] Speaker 04: A couple of the better sites for that in the appendix are computer text that discuss this at appendix pages 1302, [00:04:24] Speaker 04: to 1310 and 1816 to 17. [00:04:27] Speaker 04: But the other thing about the dictionary is what we're talking about four doesn't run both ways. [00:04:31] Speaker 04: It's true that the logical operators are not necessarily arithmetic. [00:04:34] Speaker 04: By definition, logical operators can do other things. [00:04:37] Speaker 04: But arithmetic operators are comprised of logical operators. [00:04:40] Speaker 01: What's your best site in the appendix for that? [00:04:44] Speaker 04: That arithmetic operators use logical operators. [00:04:46] Speaker 04: The two that I just gave you are pretty good. [00:04:48] Speaker 04: They're computer texts that discuss [00:04:50] Speaker 04: how arithmetic operators work and show that they use logical operators, and that's appendix pages, this text for appendix pages 1302 to 10, and 1816 to 17. [00:05:01] Speaker 04: I also didn't know that there was really a dispute on that. [00:05:04] Speaker 04: In other words, I think their argument, as I understood it, was that, from their brief, was that they say that, well, we're looking at all wrong. [00:05:12] Speaker 04: We're looking at things at the bits and bytes level. [00:05:14] Speaker 04: And here, arithmetic operators are different because they're combinations of logical operators. [00:05:19] Speaker 04: Now, that argument just doesn't work for a couple reasons. [00:05:22] Speaker 01: I guess their position is that these things are categorized differently. [00:05:26] Speaker 01: There's arithmetic operators, and there's logic operators. [00:05:30] Speaker 04: Right, and there's a chart that shows them broken up differently. [00:05:31] Speaker 01: More than ordinary skeleton, yes. [00:05:33] Speaker 04: And that's their third piece of evidence, the chart that shows arithmetic operators, logical operators. [00:05:37] Speaker 04: But again, it's not inconsistent with our view of reality, because it's true that logical operators not necessarily arithmetic. [00:05:45] Speaker 04: Arithmetic operators are law operators that do particular things, but arithmetic operators are comprised of the basic building blocks, the logical operators. [00:05:52] Speaker 02: To the extent you're using on-off states, zeros and ones, in order to do anything, you have to have some sort of logical operation. [00:06:02] Speaker 04: Yes. [00:06:03] Speaker 02: Which is why one... And which includes, in order to do anything, includes then writing a program which says one plus one equals two in an arithmetic fashion. [00:06:16] Speaker 04: I take it the point may be that, OK, the claim instruction is just awfully broad. [00:06:22] Speaker 04: And at that point, is there a lot left of the claim instruction? [00:06:26] Speaker 02: No, my point is I don't see how it can be any other way. [00:06:32] Speaker 04: That's absolutely right. [00:06:33] Speaker 04: It's basic principle of digital computing is that you're going to be using logical operators if you are manipulating binary data at the bit level of the claim instruction, which is why we have to be right. [00:06:41] Speaker 04: And I think Buxton's alpha blending is the best example of that. [00:06:45] Speaker 04: It also bears emphasis on alpha blending, that they've accused the exact same alpha blending equation of infringement in the underlying litigation. [00:06:53] Speaker 04: And, I mean, here, even in their brief in this court, page 73, footnote 2, they say that the, they're telling this court that the exact same alpha blending equation, quote, relied on logical operators to perform the accused blending operation. [00:07:05] Speaker 04: That's their position for purposes of infringement, that they're repeating in this court, and it's the exact same alpha blending equation that's at issue here. [00:07:13] Speaker 04: And at a minimum, [00:07:14] Speaker 04: That's pretty good evidence of how a person of skill in the art understands the alpha blending equation. [00:07:19] Speaker 04: In the other litigation, their expert also testified in his deposition that the way that they contended the... Was this information before the board? [00:07:28] Speaker 04: Absolutely. [00:07:30] Speaker 04: And the board said that, well, there's no judicial estoppel on this last point. [00:07:33] Speaker 04: And that's true. [00:07:34] Speaker 04: The technical requirements for judicial estoppel are not satisfied. [00:07:37] Speaker 04: But there's a couple things about it. [00:07:39] Speaker 04: One, now they're telling this court that that same equation infringes. [00:07:44] Speaker 04: they're right. [00:07:45] Speaker 04: And three, it certainly goes to the perspective of a person skilled in the art, as does the deposition testimony, which is in the appendix, the inventor deposition testimony appendix pages 1485 to 88, where he repeatedly. [00:07:57] Speaker 04: What was the page number? [00:07:59] Speaker 04: Sorry, it's 1485 to 88. [00:08:01] Speaker 04: The inventor repeatedly testifies over and over and over that alpha blending is how they envisioned the asserted invention to work. [00:08:11] Speaker 04: And so for those reasons, that's how [00:08:15] Speaker 04: it's logical that that's how a person of skill in the art would read. [00:08:18] Speaker 04: And also, as a matter of the broadest reasonable interpretation standard, the broadest reasonable interpretation ought to at least be broad enough to encompass infringement theory. [00:08:26] Speaker 04: And the PTAB's actually found that in a presidential decision last year, entitled Exportation All-Housing. [00:08:34] Speaker 04: But that's why, the other, that's the basics of alpha-blending, is how it, alpha-blending is arithmetic, which is necessarily logical. [00:08:42] Speaker 04: And generally speaking, also the claim construction is just manipulating binary values at the bit level. [00:08:48] Speaker 04: And even their chart that they have in their brief of arithmetic operations and logical operations, the multiplication arithmetic operations expressly say that they're dealing with 64 bits. [00:09:00] Speaker 04: So again, manipulating binary value at the bit level, which of course is what digital computers do. [00:09:06] Speaker 02: Is XORing Boolean? [00:09:09] Speaker 04: Yes. [00:09:10] Speaker 04: That's the exclusive OR. [00:09:11] Speaker 04: And that is a Boolean logical operator that's also expressly set forth. [00:09:15] Speaker 04: And it's a separate theory about Buxton. [00:09:18] Speaker 04: It's also expressly set forth there. [00:09:21] Speaker 04: There's no question that's a Boolean logical operator. [00:09:23] Speaker 04: Their argument on that is that, OK, that's a disclosure. [00:09:26] Speaker 04: And this is separate and apart from alpha blending. [00:09:28] Speaker 04: Their argument there is, OK, there's an expressed disclosure of the Boolean logical operator there. [00:09:33] Speaker 04: But it's not used in a way that would satisfy surrounding claim limitations, which is a series of second level concerns that the board did not reach. [00:09:41] Speaker 04: Therefore, at a minimum, we would say that on the XOR, that would be at least a remand for the additional considerations that they want to raise there. [00:09:50] Speaker 04: But on AlphaBlend, we don't frankly see a need for remand, because on AlphaBlending, the record is what it is. [00:09:56] Speaker 04: The facts all line up in the same way. [00:09:58] Speaker 04: And the point is just that combinations of logical operators satisfy the claim limitation, which, by the way, the specification expressly says that logical operators can be combined, which in our view is the whole [00:10:10] Speaker 04: ballgame with respect then to arithmetic operators. [00:10:13] Speaker 04: I'll save the remainder of my time for a bottle. [00:10:16] Speaker 03: We will save it for you, Mr. Joseph or Mr. Babcock. [00:10:25] Speaker 00: Good morning, Your Honor. [00:10:25] Speaker 00: May I please the court? [00:10:26] Speaker 02: Do you agree that XOR operator is a Boolean logical operator? [00:10:30] Speaker 00: It is not used for blending, Your Honor. [00:10:32] Speaker 00: Yes, it's a Boolean operator, but it's not used for blending. [00:10:35] Speaker 00: The basic questions here were addressed by the board [00:10:39] Speaker 00: The question here, raised by Google, is whether or not arithmetic operation, no dispute, alpha blending, arithmetic operation, is that a logical operation? [00:10:51] Speaker 01: Or does it include logical operations? [00:10:55] Speaker 00: Correct. [00:10:57] Speaker 00: Does it meet the limitation that's recited of doing blending using a logical operation, using logical operators? [00:11:07] Speaker 00: Page 24 of the board's opinion, appendix 24, the board had all of the testimony from both experts. [00:11:14] Speaker 00: Google's expert said the building block theory. [00:11:19] Speaker 00: The building block theory doesn't answer the question. [00:11:23] Speaker 00: Does alpha blending manipulate bits at the binary level? [00:11:32] Speaker 00: the process of mathematical or in this case the alpha blending which uses rather logical operators, it uses arithmetic operators, does it manipulate bits at the binary level? [00:11:45] Speaker 00: And the answer is no, it does not. [00:11:48] Speaker 00: The alpha blending and the arithmetic. [00:11:50] Speaker 02: Does it utilize them at the binary level? [00:11:53] Speaker 00: No, not the coding. [00:11:54] Speaker 00: No? [00:11:55] Speaker 02: So the computer doesn't work? [00:11:56] Speaker 02: Pardon me? [00:11:57] Speaker 02: So the computer doesn't work if it's not utilizing them at the binary level. [00:12:01] Speaker 00: The arithmetic operators, ultimately, at some level of abstraction, down at the very basic level of computing, they use bits and bytes. [00:12:10] Speaker 00: But arithmetic operators do not manipulate bits at the binary level. [00:12:15] Speaker 00: They manipulate bytes, possibly larger groups of bits, but they don't manipulate bits. [00:12:20] Speaker 02: The advantage of using... But do they utilize them? [00:12:22] Speaker 02: I'm not asking if they manipulate them. [00:12:24] Speaker 02: Do they utilize them? [00:12:25] Speaker 02: Are they part of the arithmetic operators? [00:12:28] Speaker 00: The question is whether or not [00:12:31] Speaker 00: They manipulate binary values at the bit level. [00:12:33] Speaker 00: That is the claim construction of the board. [00:12:36] Speaker 00: Do they manipulate binary values at the bit level? [00:12:41] Speaker 00: Not utilized, but do they manipulate? [00:12:43] Speaker 00: The board says, we looked at all the evidence. [00:12:47] Speaker 00: We looked at all the arguments. [00:12:48] Speaker 00: We agree with patent owners. [00:12:50] Speaker 00: We determine that petitioners' arguments do not provide us that Buxton's alpha blending equation uses operators that manipulate binary values at the bit level. [00:13:00] Speaker 00: consistent with our construction of logical operators, which is not in dispute. [00:13:04] Speaker 00: The alpha blending equation unquestionably involves arithmetic operations, which we find differ from logical operations. [00:13:13] Speaker 01: Because of the operating at the byte level versus... Exactly. [00:13:17] Speaker 00: Not because... Binary, exactly. [00:13:19] Speaker 00: Binary, using logical operators does manipulate bits at the bit level, making it more efficient. [00:13:27] Speaker 00: Alpha blending, which uses no dispute, uses [00:13:30] Speaker 00: arithmetic operations does not manipulate at the bit level. [00:13:33] Speaker 00: It manipulates at larger levels, larger blocks of data. [00:13:37] Speaker 00: Now, if you go further down into the computer, your honor, yes, there will be bits used, but they're not being manipulated at the bit level. [00:13:45] Speaker 00: That's exactly what the board found. [00:13:47] Speaker 00: That's a factual finding that is supported by substantial evidence. [00:13:51] Speaker 00: There's no evidence in this record that says alpha blending or arithmetic operations manipulate bits [00:13:59] Speaker 00: or binary values at the bit level. [00:14:03] Speaker 00: Do logical operations manipulate the arithmetic operations? [00:14:07] Speaker 00: No. [00:14:08] Speaker 00: No. [00:14:08] Speaker 00: There are two different categories of operations, as the board found. [00:14:14] Speaker 00: If you take the construction that Google is offering, this limitation means nothing. [00:14:22] Speaker 00: This limitation means the claim is a nullity. [00:14:24] Speaker 00: You have two options here. [00:14:25] Speaker 00: You could say, well, everything, every operation [00:14:28] Speaker 00: It's ultimately a bid operation. [00:14:31] Speaker 00: It doesn't meet this limitation. [00:14:32] Speaker 00: But you can just say, well, generally, it's all working with bits and bytes. [00:14:34] Speaker 00: It's all the same. [00:14:35] Speaker 00: Therefore, this limitation means nothing. [00:14:38] Speaker 00: Or you can do what the board did and say, presumably, this limitation was added to narrow the claim that it was. [00:14:43] Speaker 02: Let me ask you about what the board did. [00:14:47] Speaker 02: And the PTAB stated that it disagreed with Google's argument that gals [00:14:51] Speaker 02: I think I'm saying that right. [00:14:52] Speaker 02: G-O-U-G-A. [00:14:53] Speaker 00: Oh, we're talking about GO. [00:14:54] Speaker 02: GO. [00:14:55] Speaker 02: OK. [00:14:55] Speaker 02: GO's color lookup table discloses logical operators. [00:14:58] Speaker 02: Correct. [00:15:00] Speaker 02: Where did it explain why the color lookup table evaded logical operators and claimed one is construed? [00:15:09] Speaker 00: So let me make sure I understand your question. [00:15:13] Speaker 00: What does the color lookup table do? [00:15:14] Speaker 00: How does it work? [00:15:15] Speaker 00: Or how did the board analyze it? [00:15:17] Speaker 02: How did the board analyze it? [00:15:19] Speaker 00: That's my question. [00:15:20] Speaker 00: Well, the board considered [00:15:21] Speaker 00: The arguments and the evidence presented by Google and by IV. [00:15:26] Speaker 02: Yes, where did they explain what they did? [00:15:29] Speaker 00: In the opinion? [00:15:30] Speaker 02: Yeah. [00:15:33] Speaker 02: They said they disagreed. [00:15:35] Speaker 00: Right. [00:15:37] Speaker 00: Well, the reason they disagreed is because the description, if you look at GO, the description of the lookup table is very general. [00:15:47] Speaker 00: There's no description of how a lookup table works. [00:15:50] Speaker 00: Now Google says, well, it's going to be using logical operators. [00:15:52] Speaker 00: That isn't true. [00:15:54] Speaker 00: A lookup table takes an input. [00:15:56] Speaker 00: Another input tells the output. [00:15:57] Speaker 00: It could be one green gives you apple. [00:16:01] Speaker 00: That's what a lookup table does. [00:16:02] Speaker 00: It doesn't necessarily use logical operators. [00:16:04] Speaker 00: It's basically giving you one parameter plus another parameter. [00:16:06] Speaker 00: It gives you an answer. [00:16:07] Speaker 00: That's what a lookup table does. [00:16:09] Speaker 00: It's not using logical operators. [00:16:11] Speaker 00: It's not necessarily using arithmetic operators. [00:16:13] Speaker 00: It could be different. [00:16:14] Speaker 00: It could be anything. [00:16:16] Speaker 00: So all the board is saying is look, there's nothing in GO based on this record, based on the evidence that shows us that GO is using arithmetic operators. [00:16:24] Speaker 02: What other than if one then green equals apple? [00:16:28] Speaker 00: Right. [00:16:30] Speaker 00: Pardon me? [00:16:32] Speaker 02: What other instructions? [00:16:34] Speaker 02: You said it could be using something else. [00:16:36] Speaker 02: What? [00:16:38] Speaker 00: All you're doing there is you're taking an input and you're saying if you get this input and you have the second input, what's the answer? [00:16:44] Speaker 00: It doesn't have to be arithmetic. [00:16:45] Speaker 00: It doesn't have to be logical. [00:16:46] Speaker 00: It's just giving you the output from two given inputs or three given inputs or whatever parameters you have input. [00:16:54] Speaker 00: You're not necessarily using math. [00:16:55] Speaker 00: You're not using arithmetic. [00:16:56] Speaker 00: You're not using logic binary. [00:16:57] Speaker 00: It's just another way for a computer to manipulate numbers, to manipulate values. [00:17:04] Speaker 00: It's not using logical operators. [00:17:06] Speaker 00: It doesn't have to use any of those. [00:17:09] Speaker 00: The question here for Go is does it anticipate [00:17:13] Speaker 00: Does it disclose? [00:17:15] Speaker 02: Show me what else other than logical or arithmetic in the record it could use. [00:17:21] Speaker 00: Well, we have the table that's described in Appendix 1824. [00:17:30] Speaker 00: And if we go to 1824, there's a list of some types of operators that are used. [00:17:38] Speaker 00: This is an IEEE [00:17:43] Speaker 00: manual, MIPS manual. [00:17:45] Speaker 00: The MIPS manual is what, this is the silicon graphics manual that this patent was based on, and you see you have categories of operators, you have arithmetic, you have logical, you have data transfer, you have conditional branch, unconditional jump, and I'm not representing, that's all there are, but there are many different types of operators that people of skill in the art would understand what they mean, and so when you claim [00:18:12] Speaker 00: a logical operator to a person of skill in the art, the testimony is the evidence is that person of skill in the art would understand that means this kind of logical operator, which is not an arithmetic operator. [00:18:23] Speaker 00: So the testimony here is from people of skill in the art providing their opinion about when you see the claim, it says using a logical operator, what does that mean to a person of skill in the art? [00:18:34] Speaker 00: Not what it means to me or to a lay person or even to this panel. [00:18:37] Speaker 00: But what does it mean to a person of skill in the art? [00:18:39] Speaker 00: The evidence is it means something different [00:18:42] Speaker 00: than arithmetic. [00:18:43] Speaker 00: And it gives the claim meaning. [00:18:45] Speaker 00: If you say, well, these are all logical. [00:18:48] Speaker 00: They all manipulate bits at some level. [00:18:50] Speaker 00: Then that claim limitation has no meaning. [00:18:52] Speaker 00: And it should not. [00:18:53] Speaker 00: It means that the drafter put in language that had zero consequence. [00:18:57] Speaker 00: And the board. [00:18:58] Speaker 01: My counsel, is it true that the board didn't rely on this table that you're pointing us to right now? [00:19:04] Speaker 00: They did rely on it. [00:19:05] Speaker 01: They did rely on it? [00:19:06] Speaker 00: Yes. [00:19:06] Speaker 01: I thought that they said that they weren't going to rely on it. [00:19:10] Speaker 01: because it was submitted later, or we brought in during oral argument. [00:19:13] Speaker 01: Am I incorrect in understanding that? [00:19:17] Speaker 00: So they did read it. [00:19:18] Speaker 00: I'm looking at my co-counselor. [00:19:19] Speaker 00: I believe they looked at it. [00:19:21] Speaker 01: I have a different question for you. [00:19:22] Speaker 01: On page 24, underneath the sentence that you directed us to earlier in the board's decision, this is JA24, it says, the alpha blending equation unquestionably involves arithmetic operations which we find differ from logical operations. [00:19:40] Speaker 01: I mean, do you agree with that statement? [00:19:42] Speaker 01: Because it sounds today as if you're really not focusing on that distinction. [00:19:46] Speaker 00: No, I am. [00:19:47] Speaker 00: i i wholeheartedly agree with that to say that medic operations are different from logical operations but yet arithmetic operations include logical no i don't believe that they could either i think they're all there are either distinctly different i think that the council for google said that they're all the same or that they include i don't i don't think arithmetic operators based on the the board's definition which is not not in dispute arithmetic operators don't manipulate bits or binary values at the bit level they manipulate larger blocks of blocks of [00:20:17] Speaker 00: blocks of data. [00:20:18] Speaker 01: Okay, I thought you were saying that the particular alpha blending equation... I'm sorry, alpha blending... What you're saying, all arithmetic operators do not operate on the bit level. [00:20:29] Speaker 00: Right. [00:20:30] Speaker 00: Arithmetic operators do not. [00:20:32] Speaker 00: Maybe I'm confusing because I'm using alpha blending and arithmetic interchangeably. [00:20:37] Speaker 00: This version of arithmetic operators is multiplication, addition, and subtraction. [00:20:44] Speaker 00: It's arithmetic. [00:20:45] Speaker 00: So this version of, as Buxton discloses, it's a purely arithmetic operation, which is different than a logical operation, manipulating data at a binary level, which is what the board found. [00:21:00] Speaker 00: And the board didn't just willingly come up with this construction on its own. [00:21:05] Speaker 00: It's a construction that the parties agreed to. [00:21:07] Speaker 00: And the board analyzed all of the evidence, including these documents we're looking at here, the expert testimony, and they said, [00:21:15] Speaker 00: they look to be different. [00:21:16] Speaker 03: A person of skill in the... Do you want to address your cross-appeal? [00:21:20] Speaker 00: Yes, Your Honor. [00:21:22] Speaker 00: With regard to my cross-appeal, I'm going to point the panel to the claim. [00:21:30] Speaker 00: You can look at 19 or 26. [00:21:37] Speaker 00: These claims require two different things to happen. [00:21:43] Speaker 00: They require selecting [00:21:46] Speaker 00: pixels, and the selecting simply means you take a pixel that's a key, that's either a key or a background, and you select to display that as a key or a background. [00:21:58] Speaker 00: That limitation is the first limitation in 19. [00:22:01] Speaker 00: It says very clearly, looking at column 13, this is appendix 120, limitation A, towards the [00:22:12] Speaker 00: the bottom of A, it says the variable pixel control causing pixels selected to form the representation of at least one key. [00:22:21] Speaker 00: Okay, so you have pixels that are selected to form a key, activated at the same time simultaneously with pixels selected to form the main image. [00:22:31] Speaker 00: And by simultaneously, it means instead of, like the prior art, you alternate one or the other. [00:22:36] Speaker 00: You're saying you take a key pixel and you activate that at the same time as you activate [00:22:42] Speaker 00: a background image pixel. [00:22:44] Speaker 00: So you activate those both at the same time. [00:22:47] Speaker 00: And what they're doing is saying, okay, key, background, key, background, key, background, and you kind of get this composite image that's composed of key pixels and main image pixels. [00:22:58] Speaker 00: But then they had to add this limitation at the end in order to get it allowed. [00:23:02] Speaker 00: The limitation at the end says wherein the variable pixel control allows individual pixels to be dedicated simultaneously to both the main image [00:23:11] Speaker 00: and the representation of at least one key. [00:23:13] Speaker 00: That's blending. [00:23:14] Speaker 00: That's a different concept. [00:23:16] Speaker 00: That's the idea of saying you take one pixel and you have it blend both key and background in one pixel. [00:23:23] Speaker 00: The patent discussion, starting with column four through five, describes those two very distinct concepts. [00:23:32] Speaker 00: Pixel selection, selecting one pixel for background or key, and displaying them simultaneously versus blending. [00:23:39] Speaker 03: Council, do you want to save any rebuttal time on the process? [00:23:42] Speaker 01: Could I ask one question? [00:23:44] Speaker 01: I'd just like to ask one question. [00:23:45] Speaker 01: I'm going back to that question about the table, which I think was Exhibit 201-8, I think. [00:23:53] Speaker 01: I believe it was Exhibit 201-8. [00:23:56] Speaker 01: And then I look at page of Appendix 28, and there there's a motion to exclude certain exhibits. [00:24:04] Speaker 01: And then it says we do not rely on any of these exhibits. [00:24:07] Speaker 01: Does that include the MIPS assembly language table? [00:24:14] Speaker 00: It wasn't excluded. [00:24:16] Speaker 01: But the board says we do not rely on any of these zip-its in rendering our decision. [00:24:22] Speaker 01: So that's how I understood them. [00:24:24] Speaker 00: That's not how I understood it, Your Honor. [00:24:25] Speaker 00: This is the border plate language for the board to punt on evidential objections. [00:24:30] Speaker 00: So I believe that [00:24:33] Speaker 01: Wait, boilerplate language? [00:24:35] Speaker 01: So the board says that there's a motion to exclude certain exhibits, which include the exhibits you're putting in front of us today. [00:24:44] Speaker 01: And the board says, we do not rely on any of these exhibits in rendering our decision. [00:24:49] Speaker 01: And you're telling me that's boilerplate? [00:24:50] Speaker 00: Well, I think what they're saying here, Your Honor, is that they did not decide the motions to exclude, because they were able to reach a decision [00:25:02] Speaker 01: without relying on those exhibits. [00:25:06] Speaker 00: Well, the expert clearly did. [00:25:08] Speaker 00: The expert, our expert, clearly relied upon this, and this was part of our expert's opinion, was how do you differentiate between arithmetic and logical. [00:25:18] Speaker 00: And this is one of many dictionaries, treatises, other things that we pointed to that said arithmetic and logical. [00:25:25] Speaker 02: Can your expert rely on evidence which has been excluded? [00:25:30] Speaker 00: Well, it wasn't excluded, Your Honor. [00:25:32] Speaker 00: It wasn't excluded. [00:25:35] Speaker 00: Judge Stowell's point, he was saying, well, to the extent that there's a dispute, we won't rely upon that specific evidence. [00:25:42] Speaker 00: But this isn't the only piece of evidence. [00:25:43] Speaker 00: This is just one piece of evidence that the board had in order to evaluate, again, not what we think is different between arithmetic and logical, but what a person of skill in the art would understand. [00:25:56] Speaker 00: And when the claim is directed to a person of skill in the art, the board looked at the evidence and said, [00:26:02] Speaker 00: they're different operations. [00:26:04] Speaker 00: I credit the testimony and the evidence provided by IV. [00:26:08] Speaker 00: And for this panel, the question is simple. [00:26:11] Speaker 00: Is there substantial evidence to support that determination? [00:26:14] Speaker 00: Not whether you would have decided it differently. [00:26:16] Speaker 01: And we shouldn't look at this table, though. [00:26:17] Speaker 01: Pardon me? [00:26:17] Speaker 01: And we shouldn't look at the table? [00:26:18] Speaker 01: Or should we look at the table? [00:26:19] Speaker 01: It's confusing because the board didn't rely on it. [00:26:22] Speaker 00: I'm not sure on rebuttal if I can address that. [00:26:24] Speaker 00: You can answer the question, sir. [00:26:27] Speaker 00: I'm going to have to go back and look at the decision. [00:26:30] Speaker 00: I believe you can look at it because I believe it was in the record and it wasn't excluded. [00:26:34] Speaker 00: I think you can look at the board's decision and say whether or not they considered it or not, it was part of the record. [00:26:40] Speaker 00: There are other pieces of evidence that were part of the record that the board looked at and said there's a difference between arithmetic and logical operations. [00:26:47] Speaker 00: Is there substantial evidence to support the board's decision in that regard? [00:26:51] Speaker 03: Yes, there is. [00:26:52] Speaker 03: Thank you, counsel. [00:26:53] Speaker 03: We'll give you a minute of rebuttal time and then cross it here. [00:26:58] Speaker 03: Mr. Josephur has almost five minutes. [00:27:04] Speaker 04: Thank you. [00:27:07] Speaker 04: In terms of the evidence not being considered, at an absolute minimum, the reason decision-making requirements of the Administrative Procedures Act require the board to identify the defined facts, identify the basis for those fact findings, and then also draw the connection between the facts found and the ultimate conclusion. [00:27:22] Speaker 04: And so when it says it's not relying on something, and quite rightly so, that's not a basis for upholding the [00:27:28] Speaker 04: the agency's decision. [00:27:29] Speaker 04: The question is, did the agency articulate a legitimate basis for the facts found and the ultimate conclusion? [00:27:35] Speaker 04: And this is a key point running through this rather cursory decision, which goes to Judge Stoll's question and also to Judge Wellick's question, which is that, no, the board, when it said about the color lookup table that there was no express disclosure, gave no reasoning whatsoever, none. [00:27:51] Speaker 04: And that's on page A16. [00:27:54] Speaker 04: Similarly, with respect to Buxton, it was said it was not persuaded. [00:27:57] Speaker 04: which again is not, it's a conclusion, not reasoning that gets you there. [00:28:03] Speaker 04: And even with that table though, I mean the table that is not really properly before you, their point that arithmetic does not operate at the bit level, I mean that table disproves it because the table says that multiplication in two different columns there describes multiplication over in comments as 64-bit outputs, bit outputs. [00:28:25] Speaker 04: Similarly, [00:28:27] Speaker 04: One of the materials I referred you to earlier describes addition. [00:28:33] Speaker 04: Sorry, I lost my page number. [00:28:35] Speaker 04: Describes, I'll get it for you, describes addition as, again, in bits in, bits out. [00:28:41] Speaker 04: And so the record shows the arithmetic does operate at the bit level. [00:28:44] Speaker 04: But even if it did not, even if we're talking about a higher level of distraction, the key point is, as I mentioned before, the specification expressly says that logical operations, quote, can be combined. [00:28:55] Speaker 04: that we should have combinations of logical operators. [00:28:59] Speaker 04: And the claim language certainly doesn't preclude that in any sense. [00:29:04] Speaker 04: Their main argument, or at least the only argument, their other argument seems to be the look, under our view, the logical operator's limitation does not really add much, because by definition, digital computers manipulate data with logical operators, and therefore we're giving meaning to this claim limit. [00:29:23] Speaker 04: The answer is yes, because logical operators do have to be used. [00:29:26] Speaker 04: But it is not the case that every word in the claim has to have particularly incremental importance with respect to the overall claim scope. [00:29:34] Speaker 04: And the important thing to remember here is, looking back at the prosecution history, this language came in when they were amending multiple independent claims to add the requirement that there be a capability for individual pixels to be blended. [00:29:45] Speaker 04: That was the patentable distinction, the capability for individual pixels to be blended. [00:29:49] Speaker 04: When they added that to three different independent claims, [00:29:52] Speaker 04: They phrased it differently in every one of them. [00:29:54] Speaker 04: This is the only one that uses the phrase logical operators. [00:29:57] Speaker 04: So that's not, that was not the thrust of the amendment. [00:29:59] Speaker 04: You could just as easily have, you know, means for blending, which by the way is what one of the other independent claims that was simultaneously amended basically says. [00:30:07] Speaker 04: But, but it has meaning in the sense that it has to be done this way. [00:30:10] Speaker 04: It's just that that's how modern digital computers work. [00:30:14] Speaker 04: On the cross appeal, we're a little confused. [00:30:20] Speaker 04: But the basic point seems to be from the reply that they think that the final composite image has to have some pixels that are unblended. [00:30:29] Speaker 04: They call that selected pixels, because they have to be in the final composite image. [00:30:31] Speaker 04: They have to have some unblended pixels because those are selected. [00:30:34] Speaker 04: And there then has to be a capability for other pixels to be, to be blended. [00:30:39] Speaker 04: So you need some unblended and a capability for some blended, I think is what they're arguing. [00:30:43] Speaker 04: The second of those is totally undisputed. [00:30:45] Speaker 04: There's no question there has to be a capability for blended pixels. [00:30:48] Speaker 04: And also everyone agrees that both of these relevant prior references disclose blended pixels. [00:30:53] Speaker 04: So the whole action seems to be on whether the final composite image has to contain some unblended pixels, which is what they call selected pixels. [00:31:00] Speaker 04: Wrong for a couple of reasons. [00:31:02] Speaker 04: First, all the claims require is, what the claims require is that pixels be selected to form the one image, that pixels be selected to form the other image, and that they're then combined in the composite image. [00:31:12] Speaker 04: That doesn't mean that the same pixel has to be selected for only one image or the other. [00:31:18] Speaker 04: There's nothing that says that at least one pixel has to be selected for only the one image or the other. [00:31:22] Speaker 04: And the specification confirms that because one of the two disclosed embodiments uses only blended pixels. [00:31:28] Speaker 04: Every pixel is selected for both images, and then the composite includes only blended pixels. [00:31:34] Speaker 04: That's spot on one of the two disclosed embodiments. [00:31:37] Speaker 04: So there's no requirement that ultimately you have any unblended pixels. [00:31:43] Speaker 04: Also, the way they're trying to get there is by saying that selected means selected for only one and not the other, essentially. [00:31:50] Speaker 04: Certainly not the ordinary meaning of selected, just chosen, not necessarily chosen for only one or the other. [00:31:55] Speaker 04: And it's essentially an unpreserved claim construction argument, because there's certainly no claim construction that selected takes that extremely unusual meaning. [00:32:04] Speaker 04: The, slightly over, if you want, I can make one more point on this or sit down. [00:32:08] Speaker 03: Thank you, Mr. Joseph. [00:32:10] Speaker 03: Thank you. [00:32:13] Speaker 03: Mr. Babcock has a minute on the cross field. [00:32:16] Speaker 00: Yes, Your Honor, thank you. [00:32:19] Speaker 00: Your Honor, you'll notice that Google's counsel didn't point you to the claims. [00:32:23] Speaker 00: Google's counsel wants to say, look, selecting, blending, it's all making one image. [00:32:31] Speaker 00: When you reach your decision on this, I want you to look at the claims. [00:32:34] Speaker 00: The claims are very clear. [00:32:35] Speaker 00: I look at claim 26 this time. [00:32:39] Speaker 00: It says pixels are selected to form the main image such that the main image and the representation of at least one input zone are displayed simultaneously. [00:32:53] Speaker 00: 19 is probably clear. [00:32:57] Speaker 00: It says the variable pixel control causing pixels selected to form the representation of at least one key. [00:33:06] Speaker 00: You're selecting pixels to do what? [00:33:08] Speaker 00: to form the representation of a key in what? [00:33:11] Speaker 00: In this composite image. [00:33:13] Speaker 00: So it's the representation of a key in the composite image. [00:33:16] Speaker 02: Did the PTAB base its construction on entirely intrinsic evidence, or was there extrinsic? [00:33:26] Speaker 00: I don't believe there was extrinsic here that the PTAB was looking at. [00:33:29] Speaker 00: I think they just looked at this very, and they said, look, we think [00:33:33] Speaker 00: simply blending pixels, this last limitation is sufficient. [00:33:38] Speaker 00: And they wrote out the first part of the limitation, which says you've got to select pixels to form the representation of a key. [00:33:46] Speaker 01: I understand them to be saying that the claims don't cite unblended pixels, and that none of these steps require that. [00:33:55] Speaker 00: But if you look at the description... I should say both blended and unblended. [00:34:00] Speaker 00: That's what the board says. [00:34:01] Speaker 00: The board says that they'll require both. [00:34:03] Speaker 00: But they read out the notion, the very explicit language that says you have to select a pixel to form the representation of a key. [00:34:13] Speaker 00: And you have to display that to select a pixel to form the representation of a main image. [00:34:19] Speaker 00: A pixel that's blended isn't the representation of a key, and it's not the representation of a main image. [00:34:25] Speaker 00: It's both. [00:34:26] Speaker 00: That's in the last limitation. [00:34:28] Speaker 00: A pixel that is formed to select the representation of a key [00:34:32] Speaker 01: does what the claim says it represents a key or it's just a minute part as I understand the last step you're referring to the wherein clause that's what you're referring to that's that's the blending that's the blending I think that's the wherein that occurs from the steps A and B that's how one would normally read the claim right? [00:34:49] Speaker 00: No it's not how one would normally read the claim. [00:34:50] Speaker 00: In fact the claim was originally drafted without the wherein clause and the examiner read that claim on devices that simply selected pixels and in order [00:35:00] Speaker 00: to get the claim allowed, you had to add an additional step, which is you have to also have the ability to blend along with the ability to select. [00:35:10] Speaker 03: Thank you, Your Honor. [00:35:11] Speaker 03: Thank you, counsel. [00:35:11] Speaker 03: We'll take the case under review.