[00:00:01] Speaker 05:
Our next case is number 18-2200, Veritas Technologies LLC versus Real-Time Data LLC.

[00:00:09] Speaker 05:
Again, Mr. Gardens.

[00:00:45] Speaker 03:
Okay.

[00:00:47] Speaker 03:
Thank you, Your Honor.

[00:00:48] Speaker 03:
May it please the Court.

[00:00:50] Speaker 01:
It seems to me, Mr. Garza, your principal problem here is that your expert testified against you.

[00:00:56] Speaker 03:
So the expert testified to what he would specifically do if he had to implement the system and combine them at his skill level and actually do it.

[00:01:05] Speaker 03:
And that's not to test for obviousness.

[00:01:07] Speaker 03:
And that was done over objection because it's not to test for obviousness.

[00:01:11] Speaker 03:
The test for obviousness is based on the teachings themselves.

[00:01:14] Speaker 03:
And based on the teachings themselves, here, it is obvious.

[00:01:18] Speaker 03:
As this court said in Allied erecting, you have to look at those teachings, not whether you can physically combine it.

[00:01:23] Speaker 03:
And indeed, in Allied erecting, it was even more complex because you had an immovable arm in the caterpillar reference.

[00:01:29] Speaker 03:
And then the Ogawa reference had two movable arms, but they were somewhat incompatible, as the patentee said.

[00:01:35] Speaker 03:
And this court said, no, that incompatibility aside, the teachings are clear.

[00:01:40] Speaker 03:
And the teachings here are clear.

[00:01:42] Speaker 03:
Could you alternatively do it where you have a data descriptor and you're lacking the data descriptor and then you stop?

[00:01:47] Speaker 03:
Yes.

[00:01:47] Speaker 03:
That's exactly what was said in the petition.

[00:01:50] Speaker 03:
Could you do it the way Dr. Crusair said if he was forced to actually build it the way he was built it?

[00:01:54] Speaker 03:
Yes, that also is obvious.

[00:01:56] Speaker 03:
And if you did it the way Dr. Crusair said to build it, this way of incorporating the data type indicator into step into module 205 of Franchek, so it goes into 401, that block at the top of 4A.

[00:02:10] Speaker 03:
If you do it that way, you're still going to have the problem of unknown data types.

[00:02:15] Speaker 05:
tell you how I see this, and you can tell me whether that's correct or not.

[00:02:20] Speaker 05:
That what you have here is a situation in which you've got Framizak, which tells you to choose a compression algorithm based on data type, based on the descriptor.

[00:02:36] Speaker 05:
and that SU tells you to do more than look at the descriptor.

[00:02:44] Speaker 05:
It's a more complicated process that helps you to choose the right algorithm.

[00:02:49] Speaker 05:
And what the board is saying is we didn't have any testimony that if you combined SU with FRANASAC that you would stop with the category of

[00:03:03] Speaker 05:
compression algorithms where there's a, or data where there's a descriptor.

[00:03:10] Speaker 05:
You would go beyond that and include all of the data and you would not need a default algorithm.

[00:03:18] Speaker 05:
So the question is, is it appropriate to say, well, if you were to combine them, the best objective would be to completely wipe out the default.

[00:03:29] Speaker 05:
Or is it also a proper, obvious analysis to say we'll use Sue to further refine the algorithm where we have a descriptor?

[00:03:44] Speaker 05:
Is that a fair description of what the issue is here?

[00:03:47] Speaker 03:
Yes, although I would disagree that there's no evidence that you'd do it, because what the evidence is that you'd stop a data descriptor is common sense.

[00:03:54] Speaker 03:
you'd stop a data descriptor because you're not going to waste time and potentially create an unknown data.

[00:03:59] Speaker 05:
There's not testimony to that effect, right?

[00:04:02] Speaker 03:
Dr. Crusair did not specifically testify to that, no, Your Honor.

[00:04:05] Speaker 05:
Okay, so let's put that aside then.

[00:04:08] Speaker 05:
So it seems to me the question is,

[00:04:13] Speaker 05:
Is it appropriate for the board to say someone would have gone all the way and eliminated the default algorithm, or can you make a combination which stops where you have a descriptor?

[00:04:29] Speaker 03:
No, and for several reasons, Your Honor.

[00:04:32] Speaker 03:
The teachings of the prior art are clear that you should retain the default prompt for unknown data types.

[00:04:37] Speaker 03:
That's clear in Franicek.

[00:04:38] Speaker 03:
That's exactly what Franicek teaches.

[00:04:40] Speaker 03:
And Sebastian teaches the same thing.

[00:04:42] Speaker 03:
So what the board did is flip the inquiry on its head.

[00:04:46] Speaker 03:
The prior art said keep the default prompt.

[00:04:49] Speaker 05:
What I'd like you to do, let's just assume that the board is correct, that there's no testimony here that tells you to stop when you go beyond the descriptor situation.

[00:05:01] Speaker 05:
Let's assume that the board is correct about that.

[00:05:04] Speaker 05:
As a matter of law, does that defeat a finding of obviousness, or is it sufficient to find obviousness that you could make a combination that would only go so far as the descriptor?

[00:05:16] Speaker 05:
And that might not be the best solution, but it is a solution.

[00:05:21] Speaker 03:
It is a matter of law that this court can reverse and hold that and say that the board is wrong because the teachings are clear in the prior art.

[00:05:30] Speaker 03:
In KSR, the Supreme Court did not rely on the expert's testimony.

[00:05:34] Speaker 03:
It looked at the prior art and it said, the test here is very clear.

[00:05:38] Speaker 03:
There they had four prior art references and the Supreme Court followed a path from the initial one that said you needed a sensor down to where you put it on the pedal to where you put it not on the pedal foot.

[00:05:50] Speaker 03:
and then you put it on the immovable portion.

[00:05:53] Speaker 03:
That's the teachings in the prior art, and the teachings here are equally clear, testimony aside or not, because the only thing that's left here, even if you assume

[00:06:04] Speaker 03:
that what Dr. Crusair testified is the correct test, and we don't agree with that, but it's a correct test to look for that bottling corporation.

[00:06:11] Speaker 03:
Even if you assume that you're going to move past the data descriptor, you're still going to have the problem of unknown data types.

[00:06:17] Speaker 03:
So is it obvious based on that teaching, that common sense principle that's already in the prior art, is it obvious to say you should still retain that default compression safety net?

[00:06:29] Speaker 03:
if you're moving past the data descriptor.

[00:06:31] Speaker 03:
And the answer to that is yes.

[00:06:32] Speaker 03:
You need a default safety net.

[00:06:35] Speaker 03:
No one of ordinary skill in the art would design it without the default safety net because Sebastian and Franicek taught you to do that and one of ordinary skill in the art.

[00:06:44] Speaker 05:
I think in making that argument you're not accepting my hypothetical.

[00:06:49] Speaker 05:
that the board is right, that someone seeking the best solution would eliminate the default by using Sue beyond the situation in which there's a descriptor.

[00:07:00] Speaker 05:
Let's assume that that's correct, which is a finding of the board.

[00:07:05] Speaker 05:
I'm not sure that that defeats obviousness because there is still a motivation to do it, to use Sue with respect to the descriptor situation and the fact that it might have been better

[00:07:17] Speaker 05:
to go all the way, it seems under some of the cases is not sufficient to defeat obviousness.

[00:07:25] Speaker 05:
In other words, it's obvious if a combination

[00:07:29] Speaker 05:
would give you the result, even though a better combination would not give you the result.

[00:07:34] Speaker 03:
Yes, Your Honor.

[00:07:35] Speaker 03:
I agree 100%.

[00:07:36] Speaker 03:
For example, in allied erecting, that was one of the arguments that the patentee made about the optimal combination.

[00:07:41] Speaker 03:
And this court rejected that.

[00:07:43] Speaker 03:
The optimal combination is not the only obvious combination.

[00:07:46] Speaker 03:
Because again, the test is not what the one particular combination has to be.

[00:07:51] Speaker 03:
As the board said here repeatedly, for example, in Apex 29, it has to be that specific combination.

[00:07:57] Speaker 03:
There's only one that's going to be created.

[00:07:59] Speaker 03:
That's legal error.

[00:08:00] Speaker 03:
That's not the test.

[00:08:01] Speaker 03:
The test is what the invention, whether the invention is within the grasp of an ordinary skilled artist and based on creativity, common sense and skill in light of the teachings of the prior art.

[00:08:11] Speaker 03:
And here it is clear, because even if you move past the data descriptor, even in that system, you would still have motivation based on common sense.

[00:08:20] Speaker 03:
And therefore, it would be obvious to do it.

[00:08:23] Speaker 03:
Motivation to stop at the data descriptor number one.

[00:08:26] Speaker 03:
If you move past it, you still have the same problem identified in Sebastian and identified in Frenicek of unknown data types.

[00:08:34] Speaker 03:
Indeed, Dr. Zegar even admitted at Apex 4091 on the record that if you have data that's random, that's past what the table on 2176 of Sue, Apex 2176,

[00:08:49] Speaker 03:
of Sue shows there's 10 categories of data.

[00:08:52] Speaker 03:
What would you do with random data?

[00:08:53] Speaker 03:
And Dr. Zegar said, well, you could create a random data category.

[00:08:59] Speaker 03:
That's unknown data category.

[00:09:01] Speaker 03:
Even he admitted that when you would have to actually sit down and deal with unknown data, you could modify Sue to have that type of default.

[00:09:09] Speaker 03:
And that's the same teaching that Franchek and Sebastian had expressedly.

[00:09:13] Speaker 03:
Franchek and Sebastian expressly taught you're going to keep the default because you're never going to have all the known data types.

[00:09:21] Speaker 02:
In passing, a couple of minutes ago, you mentioned Baudelaire Corporation.

[00:09:26] Speaker 02:
Yes, Your Honor.

[00:09:26] Speaker 02:
Is it still your position?

[00:09:28] Speaker 02:
that the board erroneously required bodily incorporation of soot.

[00:09:33] Speaker 03:
Yes, Your Honor, for one specific reason.

[00:09:35] Speaker 03:
On Apex 29, they looked for that one particular combination of the prior art and how it would function together.

[00:09:41] Speaker 03:
And so they said it's incompatible.

[00:09:43] Speaker 02:
And as this Court recognizes- But they required that you prove that a person of skill would have been motivated to combine frangiac and soot to satisfy the claim limitations.

[00:09:53] Speaker 02:
Yes, sir.

[00:09:53] Speaker 02:
It's not the same thing.

[00:09:55] Speaker 03:
That's not the same thing, but what they did is they went a step further.

[00:09:59] Speaker 03:
And they said all the requirements, all the criticisms in active video that this court said of Verizon's expert are prerequisites, that you have to show how one particular, as they put it, this specific, one specific particular combination would actually work.

[00:10:13] Speaker 03:
That's not the test for obviousness, as this court has made clear time and time again.

[00:10:16] Speaker 05:
The teaching...

[00:10:18] Speaker 05:
I'm not really following what you're saying.

[00:10:21] Speaker 05:
It seems to me what the board is saying is that if you combine Sue with Frantzak, you will have used Sue to such an extent that there will be no longer a need for a default algorithm.

[00:10:35] Speaker 05:
And your witness seemed to say, yes, that might be the best solution.

[00:10:41] Speaker 05:
Does a showing of obviousness require that when you make the combination that you go whole hog and achieve the best solution, or is it sufficient that you could stop here with the descriptor category and achieve the patented invention?

[00:11:02] Speaker 03:
Yes, Your Honor, and this court's case law says that.

[00:11:04] Speaker 03:
It doesn't have to be optimal.

[00:11:06] Speaker 03:
Even inferior alternatives, as this court has said, inferior alternatives that are still obvious can invalidate.

[00:11:11] Speaker 03:
And that's exactly what we have here, assuming that paradigm.

[00:11:15] Speaker 03:
The inferior alternative potentially would be to stop a data descriptor.

[00:11:18] Speaker 03:
Is there a reason, a motivation to do that?

[00:11:20] Speaker 03:
Yes, the motivation is clear.

[00:11:21] Speaker 03:
It's common sense.

[00:11:22] Speaker 03:
You'd stop because you're not going to waste time finding something that you know is going to slow down the system.

[00:11:27] Speaker 03:
You're going to go there, and you're going to just use a data descriptor.

[00:11:30] Speaker 03:
But importantly, what the board ignored, what they said is you're going to incorporate Sue.

[00:11:36] Speaker 03:
If you recall the chart we have in our reply brief, what they said is you're going to incorporate Sue at both prongs of the Franchet.

[00:11:44] Speaker 03:
And that's not what Dr. Crusair was saying.

[00:11:46] Speaker 03:
He was saying you'd use it at 401.

[00:11:49] Speaker 03:
You'd incorporate data type into 205.

[00:11:52] Speaker 03:
and therefore you'd choose it better.

[00:11:54] Speaker 03:
He did not say you would incorporate it in Toto and therefore you do not need default.

[00:12:00] Speaker 03:
In his deposition testimony, he did not talk about the default wing because it's clear from the prior art that you still need it.

[00:12:06] Speaker 03:
What Dr. Crusair did talk about is how even in Sioux, the Unix file system, you're still gonna have unknown data types.

[00:12:14] Speaker 03:
So there's still a motivation at that point, these unknown data types, to include a default compressor, and that's obvious.

[00:12:23] Speaker 05:
Okay, want to save your rebuttal time?

[00:12:25] Speaker 05:
Yes, unless there's questions.

[00:12:27] Speaker 05:
Thank you.

[00:12:29] Speaker 05:
Mr. Neruzzi.

[00:12:50] Speaker 05:
Thank you, Your Honor.

[00:12:52] Speaker 05:
So I'm just wondering whether the board's approach here is consistent with our case.

[00:12:57] Speaker 05:
Assuming that it's correct, that someone skilled in the art at combining friend Zach and Sue would say, well, the best solution here is to use Sue all the time, not just in the descriptor situation.

[00:13:12] Speaker 05:
And that would make it unnecessary to have a default algorithm.

[00:13:15] Speaker 05:
Whereas another solution, maybe not as optimal,

[00:13:19] Speaker 05:
But another solution would be to use Sue only in the descriptor situation and retain the default algorithm.

[00:13:27] Speaker 05:
And our cases seem to be pretty clear that in an obviousness situation that you don't have to take the combination all the way.

[00:13:37] Speaker 05:
if there's a motivation to replace part of the prior art and that's going to render it obvious even though the optimal solution would be to go all the way.

[00:13:50] Speaker 05:
So comment on that.

[00:13:52] Speaker 04:
Yes, thank you, your honor.

[00:13:53] Speaker 04:
So I actually agree with everything you said, and I think it supports the board's decision.

[00:13:59] Speaker 04:
And the reason is the following.

[00:14:01] Speaker 04:
It's the part of your statement where you said if there is a motivation to do the other approach that may not be the best.

[00:14:08] Speaker 04:
the optimal approach.

[00:14:09] Speaker 04:
And that's what's lacking here.

[00:14:11] Speaker 05:
So the board did not find... Well, there certainly is a motivation to use Su to get a more detailed study of the data to choose the right algorithm.

[00:14:26] Speaker 05:
And I don't see that the new witness disagreed with that, right?

[00:14:31] Speaker 04:
Well, first of all, our witness did disagree, but he did say that, and this is the key point, that if you do use SUE to find a better understanding of what the data type is when Franchek only has a descriptor,

[00:14:46] Speaker 04:
then you certainly would keep using Sue to find out what the data type is if Franchek does not have a descriptor, because finding out what the data type is lets you get a better compression algorithm.

[00:14:57] Speaker 04:
So why would you not want to know what the data type is?

[00:15:00] Speaker 04:
Well, that all may be true, but that's the optimization.

[00:15:04] Speaker 05:
theory that if you make the combination, you have to choose the best solution rather than a good solution.

[00:15:14] Speaker 04:
No, Your Honor.

[00:15:15] Speaker 04:
It's not an optimization theory, respectfully.

[00:15:18] Speaker 04:
What's going on here is that the petitioner, now the appellant, had a very specific theory that you do use Sue when there is a data type, and then you don't use Sue

[00:15:29] Speaker 04:
when there's no data type.

[00:15:30] Speaker 04:
That's their theory, and that's what we're talking about, whether someone would be motivated to do both of those things.

[00:15:36] Speaker 04:
And the board said, well, you need to prove both things for your motivation to combine, not just the first one.

[00:15:42] Speaker 05:
Well, they said you need to stop, because there would be a motivation to make it better by using Sue in all situations.

[00:15:51] Speaker 05:
So I'm not sure that the case law supports the notion that there has to be a motivation to stop once you've applied suit in the descriptor category.

[00:16:00] Speaker 04:
Your Honor.

[00:16:01] Speaker 04:
That's not the board statement.

[00:16:04] Speaker 04:
That's the board identifying correctly what the petitioner's combination theory was, and then telling the petitioner, well, that's your theory.

[00:16:14] Speaker 04:
You need to prove it.

[00:16:16] Speaker 04:
And so as odd as the theory sounds, it was the petitioner's theory.

[00:16:21] Speaker 04:
The petitioner's theory is, and now the appellant's theory,

[00:16:25] Speaker 04:
that if you have a data type in Franchek, go ahead and use Sue to do further analysis to get a better idea of the data type.

[00:16:33] Speaker 04:
That already is a little bit odd, but let's put that aside.

[00:16:36] Speaker 04:
Then they have a further requirement in their theory, which is, but if Franchek doesn't know the data type at all, doesn't even have a descriptor,

[00:16:44] Speaker 04:
Don't use Sue to figure out what the data type is.

[00:16:47] Speaker 04:
Don't touch Sue, right?

[00:16:49] Speaker 04:
Okay, well, that's your theory.

[00:16:51] Speaker 04:
You have to prove it.

[00:16:51] Speaker 05:
And the board said that you have to prove your whole theory.

[00:16:55] Speaker 05:
I mean, you're right.

[00:16:56] Speaker 05:
The board rejected the second part of that.

[00:17:00] Speaker 05:
And the board's finding in that respect seems to be supported by the record that there wasn't any proof of a motivation to stop.

[00:17:08] Speaker 05:
But I guess what I'm asking you is why do you have to have a motivation to stop?

[00:17:14] Speaker 05:
Why isn't it sufficient that you apply suit of descriptor category and that gives you something that's better and that's sufficient?

[00:17:25] Speaker 04:
I see.

[00:17:26] Speaker 04:
Your Honor, the reason that the petitioner applied it that way and the board required them to prove it and that there has to be a motivation to stop is because that's what the limitations of the claims require.

[00:17:37] Speaker 00:
It's the last limitation to claim that you're focusing on.

[00:17:40] Speaker 04:
That's right, Your Honor.

[00:17:42] Speaker 04:
So, limitation 1F and 24F say that if you do not find any parameter attribute at all, then use this default

[00:17:54] Speaker 04:
encoder and that if that predicate if is the key thing that they need to show is also obvious to me under the priority.

[00:18:02] Speaker 04:
They can't show that lack of an identification of any parameter attribute if they don't stop using Sue because they themselves admit that Sue will always find a parameter attribute and that's in the blue brief at page

[00:18:23] Speaker 04:
17, I believe.

[00:18:25] Speaker 04:
And so, SU uses a statistical analysis approach.

[00:18:29] Speaker 04:
Statistics can't fail to find some information about a block of data, redundancy methods and so forth.

[00:18:36] Speaker 04:
Every time you run it, you'll get an answer.

[00:18:39] Speaker 04:
And those are parameters or attributes.

[00:18:41] Speaker 04:
So, if you're using SU and you don't stop, you will meet the first

[00:18:46] Speaker 04:
part of the claim, but you will never meet that last part of the claim.

[00:18:50] Speaker 04:
And so to meet that last part of the claim, they recognize this in their petition.

[00:18:54] Speaker 04:
They said, okay, now we have to additionally say there's a situation where you would stop and you would not use Sue at all.

[00:19:00] Speaker 04:
And that's pretty anomalous.

[00:19:02] Speaker 04:
If you don't know what the data type is in Franchek, and you've got this great system in Sue for finding out what the data type is, why would you want to stop?

[00:19:10] Speaker 04:
They need that for their theory.

[00:19:12] Speaker 05:
Sure, to optimize it, you would.

[00:19:13] Speaker 05:
go all the way, and both experts seem to agree with that.

[00:19:17] Speaker 05:
But I'm not sure that the case law suggests there has to be a motivation to stop, as long as there's a motivation to go at least as far as the descriptor category, which there certainly is.

[00:19:28] Speaker 04:
Well, then you don't meet the last limitation, Your Honor.

[00:19:31] Speaker 04:
That's the problem.

[00:19:32] Speaker 05:
How do you not meet the last limitation?

[00:19:33] Speaker 04:
Because to meet the last limitation of the claim, it has to be that no parameter or attribute is identified.

[00:19:42] Speaker 04:
And the only way that happens in their combination is if you do not use SUE when Franchek doesn't have a data type.

[00:19:51] Speaker 04:
Does that make sense?

[00:19:53] Speaker 04:
If Franchek does not have a data type, but you do use SUE, then you will find a parameter attribute, always.

[00:20:02] Speaker 04:
So their theory, which they set up themselves, requires the stop portion.

[00:20:08] Speaker 05:
You're suggesting that the last claim limitation means that you can't have a descriptor?

[00:20:13] Speaker 04:
No, Your Honor.

[00:20:16] Speaker 04:
The last limitation means not only can you not have a descriptor, but you can't have any parameter attribute that gives you an indication of the data type.

[00:20:25] Speaker 00:
It's something that can't be identified, right?

[00:20:27] Speaker 00:
I mean, the exact limitation says if one or more parameters or attributes of the data are not identified.

[00:20:34] Speaker 00:
And I understand, correct me if I'm wrong, that your position is that Sue would always identify the parameters or attributes of the data.

[00:20:43] Speaker 04:
and and that's not only yes your honor that is correct and that's not only our position that's also the appellant's position they say this in the blue brief and we discuss it in our brief we point out that they admit at blue brief 35 i'm sorry i correct my prior site blue brief 35 quote the redundancy metrics of sue can always be calculated because they are statistical formulas

[00:21:14] Speaker 04:
And so I'd like to directly address, Judge Dyke, your question about whether there's a need for a motivation to stop under the law.

[00:21:23] Speaker 04:
That's not the issue, respectfully, in this proceeding.

[00:21:29] Speaker 04:
In this appeal, the question is whether there's a requirement under the law to have a motivation

[00:21:36] Speaker 04:
that is adequate to create the combination in a way that will meet all of the claim limitations, not just some of them.

[00:21:43] Speaker 04:
That's the question.

[00:21:45] Speaker 04:
And their motivation to start, if we're going to simplify that way, gets them to some of the claim limitations.

[00:21:52] Speaker 04:
But for them to be able to get to the rest of the claim limitations, they need the motivation to stop.

[00:21:58] Speaker 05:
I'm still not clear on why this F limitation

[00:22:03] Speaker 05:
is inconsistent with combining SUE with Franiček and stopping with the descriptor category.

[00:22:14] Speaker 04:
Your Honor, to get there, to get there, you would only have to use SUE when Franiček has a data type.

[00:22:22] Speaker 04:
You would have to have the motivation to only do that.

[00:22:25] Speaker 04:
No, but that sounds like a motivation to stop.

[00:22:28] Speaker 04:
No, Your Honor, that's just a motivation of how you use Sue in combination with Franchek.

[00:22:35] Speaker 04:
To get to the claims as a whole, you would have to be motivated to only use Sue when Franchek has a data type and to not use Sue when Franchek lacks a data type.

[00:22:46] Speaker 04:
And their experts said, I don't think that's right.

[00:22:49] Speaker 04:
I wouldn't do that.

[00:22:49] Speaker 04:
I don't think that's motivated.

[00:22:51] Speaker 04:
I didn't opine on that.

[00:22:52] Speaker 04:
I didn't say that that's what someone would do.

[00:22:54] Speaker 04:
And our experts said, yes, people would not

[00:22:57] Speaker 04:
think that it's obvious to do that.

[00:22:58] Speaker 04:
That doesn't make sense if you don't know what the data type is.

[00:23:01] Speaker 04:
You would want to go and find out what it is if you have a good tool for it.

[00:23:05] Speaker 04:
So their theory to meet the claim requires that they provide that motivation and they don't have evidence of it.

[00:23:12] Speaker 04:
And that's the basis for the board's finding is that there is a necessary element of their combination that needs a motivation to meet the claims.

[00:23:20] Speaker 04:
And that necessary element does not have a motivation.

[00:23:31] Speaker 04:
Your honors, if there are no further questions for me, I will give back the rest of my time.

[00:23:36] Speaker 05:
OK.

[00:23:37] Speaker 05:
Thank you, Mr. Chairman.

[00:23:38] Speaker 05:
Mr. Goddard?

[00:23:51] Speaker 03:
Your Honor, you're absolutely right that that would be a motivation to stop.

[00:23:54] Speaker 03:
And as this Court has repeatedly held,

[00:23:56] Speaker 03:
In a lighter valid correcting and we moved it which we cite on page 14 of a reply brief.

[00:24:02] Speaker 03:
It does not have to be optimal and it can be an inferior Option and it would still be obvious and here in light.

[00:24:09] Speaker 00:
Can you explain how it would meet that last claim limitation?

[00:24:13] Speaker 00:
I mean, I understand as a merit law.

[00:24:15] Speaker 00:
It doesn't have to be an optimal combination

[00:24:18] Speaker 00:
But the argument being made is that if you only modify it to a certain point, you're not going to, I mean, if you modify it in view of sue, you're not going to meet the last limitation to claim.

[00:24:30] Speaker 00:
So you need to make a further modification of the combined references in order to make sure it meets the last limitation.

[00:24:38] Speaker 00:
And if you don't have a reason to modify, it might sound like hindsight.

[00:24:43] Speaker 00:
So can you address my concern?

[00:24:45] Speaker 03:
I believe I understand that, Your Honor, and it goes back to the statistical redundancies.

[00:24:50] Speaker 03:
That's what real time was talking about, is you're always going to be calculating these statistical redundancies.

[00:24:56] Speaker 03:
And that's wrong on several fronts.

[00:24:58] Speaker 03:
Number one, it doesn't make sense.

[00:25:00] Speaker 03:
If you don't have a data type,

[00:25:02] Speaker 03:
You're going to stop.

[00:25:03] Speaker 03:
You're not going to calculate statistical redundancies, and you're not going to have been calculated the data type, and the reason is in SUE itself.

[00:25:11] Speaker 03:
As the chart shows on 2176, it's a cross-reference.

[00:25:16] Speaker 03:
You need data type to ever use the statistical redundancy.

[00:25:20] Speaker 03:
Why is F in there then?

[00:25:23] Speaker 03:
Sub F. Why is it in there?

[00:25:27] Speaker 03:
It's basically a bifurcated choice.

[00:25:29] Speaker 03:
If you don't have the data attribute, then you're going to do default.

[00:25:32] Speaker 03:
That basic idea was taught by Franachek.

[00:25:35] Speaker 03:
If you don't know the data type, then you're going to do a default.

[00:25:39] Speaker 03:
The problem with Franachek supposedly is that it just has a data type indicator.

[00:25:44] Speaker 03:
So the edit to Franachek is very simple here.

[00:25:46] Speaker 03:
It's not anything that's complex.

[00:25:49] Speaker 03:
Instead of just the data type indicator, you're going to try to look for the data type of the block and potentially statistical redundancies if you find the data type.

[00:25:57] Speaker 03:
if you actually know that if you don't have the indicator, you can still do that.

[00:26:02] Speaker 03:
But going back to your question, if you don't have the indicator, you still practice and you're not going to calculate the statistical redundancies.

[00:26:09] Speaker 03:
Number one, because you're not going to do it because it's a waste of time.

[00:26:11] Speaker 03:
But number two, because there's nothing associated with it.

[00:26:13] Speaker 03:
And the claims indisputably require an associated encoder.

[00:26:17] Speaker 03:
And it's not just claim one, it's claim 24 too.

[00:26:20] Speaker 03:
Claim 24 clearly states that the encoders are associated with.

[00:26:24] Speaker 03:
And that was never argued to be any different, claim one and 24.

[00:26:27] Speaker 03:
So you have to have an encoder associated with the statistical redundancies.

[00:26:32] Speaker 03:
And in the SUE system, there's no encoder associated with the statistical redundancies.

[00:26:38] Speaker 03:
You need the data type.

[00:26:40] Speaker 03:
If you don't have the data type, you're not going to perform the content dependent compression.

[00:26:45] Speaker 03:
That's prong 1E or 24E.

[00:26:47] Speaker 03:
You're going to go to default.

[00:26:49] Speaker 03:
That's common sense.

[00:26:50] Speaker 03:
And it's right there in SUE that you need to have data type in order to even use these statistical redundancies.

[00:26:57] Speaker 03:
So what you're left with here is Frenicek that teaches the heart of the system, and you simply have to add a data descriptor.

[00:27:04] Speaker 03:
So is that within the grasp of ordinary skilled artisan creativity and based on prior art teachings?

[00:27:12] Speaker 03:
Absolutely.

[00:27:13] Speaker 03:
They're just going to add that data type.

[00:27:14] Speaker 03:
And the board reached a contrary conclusion by saying, well, your expert said that's not going to be optimal, but it doesn't have to be optimal, number one.

[00:27:21] Speaker 03:
And number two, the board reached that conclusion by saying, you're going to incorporate Sue in TOTA.

[00:27:26] Speaker 03:
And then you have to tell us how you're going to take it back.

[00:27:28] Speaker 03:
But that's not the test for obviousness either.

[00:27:30] Speaker 03:
Here, it's very clear.

[00:27:33] Speaker 03:
Thank you.