[00:00:57] Speaker 01:
Good morning.

[00:01:07] Speaker 00:
You may please the court.

[00:01:09] Speaker 00:
The claimed invention provided a new manner of emulating streaming, which improved over the prior art version of pipeline.

[00:01:16] Speaker 00:
Specifically, the prior art or conventional option for pipelining, as described in the flaw reference, used dynamic interaction between a remote server and a player in real time.

[00:01:27] Speaker 00:
The present invention avoided the need for that dynamic interaction by providing a unique control information file.

[00:01:33] Speaker 00:
Now Google's argument below in the petition was that it would take a SMILE file and modify the same to implement the invention.

[00:01:41] Speaker 00:
However, it was established that a SMILE file could not implement pipelining of any form and Google's own expert agreed.

[00:01:48] Speaker 00:
Moreover, it was established that Hua's dynamic technique would not use control information files at all and was incompatible with SMILE.

[00:01:56] Speaker 00:
And this gets to the heart of the problem.

[00:01:58] Speaker 00:
The board ultimately sidestepped these deficiencies by relying explicitly and repeatedly on the patentee's own disclosure as a solution for these problems.

[00:02:09] Speaker 00:
This was done in lieu of modifying SMILE and in lieu of combining HWAL, and that constitutes a legal error.

[00:02:17] Speaker 01:
Just to be clear, the board had two alternative holdings, one in which it relied on

[00:02:25] Speaker 01:
what it said, which was a combination of smile and wah, and the second was more or less what the petitioner was saying, was smile and general knowledge as indicated in wah, right?

[00:02:40] Speaker 01:
Correct.

[00:02:41] Speaker 01:
So the arguments you've made go more to the first or the second?

[00:02:48] Speaker 00:
They go to both, Your Honor.

[00:02:50] Speaker 00:
If you read the description in the final decision as to the first, what they're saying is we're combining, we're not even combining anything.

[00:02:57] Speaker 00:
We're having SMILE sit out there on its own because it can't do everything that's required of the invention.

[00:03:02] Speaker 00:
And then we're doing what's actually described in the specification.

[00:03:05] Speaker 00:
Now with respect to the combination of Smile and Hua, there was no argument below that Smile and Hua would be combined, specifically the conventional pipelining in Hua.

[00:03:16] Speaker 00:
Hua's conventional pipelining was this very thing that the patents said it was overcoming.

[00:03:20] Speaker 00:
It's this dynamic interaction.

[00:03:22] Speaker 00:
And in that interaction, there is no control information file.

[00:03:26] Speaker 00:
because the segments are created in real time.

[00:03:29] Speaker 00:
And there are no alternatives for a given segment, because if you're customizing your division of media, you don't need to have an alternative.

[00:03:38] Speaker 00:
You're basically making a custom-tailored item, as opposed to what the invention is, which is basically an off-the-rack solution saying, do you want a small, do you want a medium, do you want a large?

[00:03:48] Speaker 00:
Now, what the board said, sort of in one or two sentences, in the ground that it created,

[00:03:54] Speaker 00:
was that you would use the pipelining in Hua.

[00:03:57] Speaker 00:
But if you're using that pipeline, the question is how are you combining it if that version of pipelining in the prior art didn't use control information files and didn't use alternatives and used explicitly what the patent was intending to overcome.

[00:04:10] Speaker 00:
And Google seems to admit as much in its arguments on appeal concerning the deficiencies at the board.

[00:04:19] Speaker 00:
Specifically, in the red brief at 39, it said there's no need to demonstrate the obviousness of any such combination.

[00:04:26] Speaker 00:
And also on the same page that there was no need therefore to show how Hua's specific approach for dividing media files into segments would be implemented in Smile.

[00:04:34] Speaker 00:
Basically, both the board and Google seem to be avoiding the fact that there is no articulation of how you would combine Hua's technique with Smile.

[00:04:45] Speaker 00:
And in fact, what the board wants to do is just keep this all separate.

[00:04:49] Speaker 00:
This can be seen specifically in the language that the board uses.

[00:04:53] Speaker 00:
And I'm looking at the appendix of 23.

[00:04:56] Speaker 00:
In particular, a media presentation could have been divided into a sequence of segments and then... Sorry, where were you reading on 23?

[00:05:04] Speaker 01:
On the second, first full paragraph, second sentence?

[00:05:06] Speaker 00:
Correct.

[00:05:08] Speaker 00:
In particular, a media presentation could have been divided into a sequence of segments and then played back in sequence using the Seq element of SMILE.

[00:05:16] Speaker 00:
There's no dispute that a SMILE file will tell you when and how to play something, but it simply doesn't have a control information file or suggest or provide the functionality for control information file that controls downloading or the timing of downloading relative to play.

[00:05:32] Speaker 00:
So what's the board's solution?

[00:05:33] Speaker 00:
Further, during playback of one segment of the media presentation, another segment could have been downloaded simultaneously using well-known multi-thread software development tools.

[00:05:44] Speaker 00:
That discussion of multi-thread software development tools is the explanation for how to implement the invention and specification.

[00:05:51] Speaker 00:
That site in there, both the site to the patent and the site to the declaration, are sites to our own disclosure.

[00:06:00] Speaker 00:
Basically, the Bulterman site that

[00:06:03] Speaker 00:
that the board relies on, if you go there, it cites back to our patent.

[00:06:09] Speaker 00:
So the board repeatedly, four times, cites to our patent for this.

[00:06:12] Speaker 00:
And then five times cites to the Baltimore Declaration, where the cited paragraph refers back to our specification.

[00:06:19] Speaker 00:
So what the board seems to be saying is this.

[00:06:21] Speaker 00:
We have Smile, but it can't do the invention, and it doesn't have the functionality, and we're not going to modify it.

[00:06:27] Speaker 00:
So we'll just play files using Smile.

[00:06:30] Speaker 00:
How are we going to get to a control information file of the present invention?

[00:06:33] Speaker 00:
We'll just do what's said in the specification.

[00:06:35] Speaker 00:
There's two problems for that.

[00:06:37] Speaker 00:
What's described in the specification is a way to implement.

[00:06:39] Speaker 00:
They shouldn't be relying on our specification at all to inform the priority.

[00:06:45] Speaker 00:
Second, what's described with respect to those multi-threading techniques are simply that the Java language exists or the Microsoft SDK language exists and then those things can do two functions at once.

[00:06:57] Speaker 00:
They can multi-thread.

[00:06:58] Speaker 00:
They can do one thing and in parallel do something else.

[00:07:02] Speaker 02:
The patent doesn't... You're suggesting the board can't rely on your spec.

[00:07:06] Speaker 02:
But what about when you say things like at column 3, line 62 of the spec, combining multiple sequenced inputs is well understood in the industry.

[00:07:18] Speaker 00:
I'm sorry, what specific line?

[00:07:19] Speaker 00:
Column 3?

[00:07:21] Speaker 02:
Line 62.

[00:07:23] Speaker 02:
When you say something is well understood in the industry, can't the board rely on something like that in your spec?

[00:07:32] Speaker 00:
We have to understand what that's describing.

[00:07:35] Speaker 00:
Java says you can do multiple sequences at once.

[00:07:38] Speaker 00:
Just because there is a computer language that says you can do two things at once does not suggest what the two things you're going to do at once are.

[00:07:46] Speaker 02:
No, that's not right.

[00:07:48] Speaker 02:
What it says here is combining multiple sequence inputs is well understood in the industry.

[00:07:53] Speaker 02:
For example, and then it goes into the Java example.

[00:07:57] Speaker 02:
So the preceding sentence is obviously much broader in nature than just the Java example.

[00:08:03] Speaker 00:
I'm sorry, are you suggesting that the sequenced elements are pipelining?

[00:08:13] Speaker 01:
I think the response was to your suggestion that the board couldn't cite anything in the spec and use it and I think this is an example of something that was

[00:08:24] Speaker 00:
your game for the board to use when you're discussing what exists in the art and what we're both the specification I'm I believe saying and even with requests to sequenced elements it's not saying what things are being sequenced Java existed and can do sequences in row it could do multiple things at the same time but just because a programming language exists

[00:08:45] Speaker 00:
that can do sequences in a row doesn't mean it specifically teaches a control information file that does the specific sequences and the specific things that we recite in the claim convention.

[00:08:57] Speaker 00:
And I want to make sure that's clear.

[00:09:00] Speaker 00:
Basically, we need to provide in our specification some indication of how this can be implemented.

[00:09:04] Speaker 00:
There has to be enablement.

[00:09:05] Speaker 00:
And all this is explaining is that the Java language allows you to do different sequences.

[00:09:10] Speaker 00:
It allows you to put things together in certain

[00:09:12] Speaker 00:
But just because the Java language exists to do so does not suggest a cultural information file that does the specific things recited.

[00:09:20] Speaker 02:
Concerning segments, concerning alternatives, and concerning the relative timing of... Let's look at claim one and why don't you tell me what the specific sequencing things are that you're referring to that wouldn't be captured by your general statement that combining multiple sequencing inputs is well understood.

[00:09:37] Speaker 00:
Well, what it says there is that there is a control information file.

[00:09:42] Speaker 02:
Claim one.

[00:09:42] Speaker 02:
Tell me what you want me to look at and claim one.

[00:09:44] Speaker 00:
Sure.

[00:09:47] Speaker 00:
The third element, the client device parsing the control information file and based on parsing the control information.

[00:09:52] Speaker 02:
How is it the third element?

[00:09:54] Speaker 02:
Isn't that the second element?

[00:09:55] Speaker 00:
I wasn't counting the preamble.

[00:09:56] Speaker 00:
I'm sorry.

[00:09:56] Speaker 02:
You weren't counting what?

[00:09:58] Speaker 00:
The preamble.

[00:09:58] Speaker 00:
So downloading and then after that.

[00:10:00] Speaker 02:
You weren't counting the preamble?

[00:10:03] Speaker 00:
So there's a preamble and then two elements.

[00:10:05] Speaker 02:
This is the third element.

[00:10:07] Speaker 02:
The second element, not the third.

[00:10:09] Speaker 02:
The client device parsing.

[00:10:11] Speaker 00:
The client device parsing the control information file and based on the parsing the control information file doing a number of things.

[00:10:18] Speaker 00:
Identifying multiple alternative files.

[00:10:22] Speaker 00:
There's nothing with respect to our discussion in Java of identifying multiple alternative files or saying that Java teaches to download one at a specific time relative to downloading another.

[00:10:34] Speaker 00:
And to play one at a specific time relative to playing another.

[00:10:38] Speaker 00:
That's what's recited in the rest of this.

[00:10:41] Speaker 00:
And what you do is parse the control information file.

[00:10:43] Speaker 01:
I'm sorry, in the rest of this what?

[00:10:44] Speaker 01:
In the other limitations 3, 4, and 5?

[00:10:47] Speaker 00:
Correct.

[00:10:47] Speaker 00:
Determining which of the multiple alternative files to retrieve.

[00:10:50] Speaker 00:
Yes.

[00:10:50] Speaker 00:
Based on system restraints.

[00:10:52] Speaker 00:
Retrieving the determined file of the multiple ones.

[00:10:55] Speaker 00:
Begin a media presentation, and then going on with concurrent, the second to the last element, concurrent with the media presentation, retrieving a next file.

[00:11:04] Speaker 00:
But just because you can retrieve multiple things at the same time doesn't mean that Java teaches or suggests that you are doing the specific retrieving of this media segment, which is a division or alternative of something else, and downloading relative to play.

[00:11:22] Speaker 00:
It's just saying that the Java language can be used to implement that.

[00:11:26] Speaker 00:
But Java itself doesn't describe all of these features.

[00:11:30] Speaker 00:
Java is just one programming language that can be used to program these elements.

[00:11:36] Speaker 00:
I'm not sure if I'm making myself clear if there's another question in that regard.

[00:11:40] Speaker 00:
But I want to make sure we work clear on that.

[00:11:43] Speaker 00:
Because what the board is saying is, or seems to be saying, is that you have Smile over here and you're not going to modify Smile.

[00:11:48] Speaker 00:
And instead, you're going to use Java.

[00:11:50] Speaker 00:
If that's the case, where does the control information file exist?

[00:11:53] Speaker 00:
Is it in the SMILE language?

[00:11:56] Speaker 00:
Is it in Java?

[00:11:58] Speaker 00:
What the claim requires is the specific use of the control information file.

[00:12:02] Speaker 00:
And the board's not modifying SMILE, and it's using the discussion of Java for our specification, where do we actually get the claim control information file used for the same?

[00:12:15] Speaker 00:
As stated in KSR,

[00:12:21] Speaker 00:
A patent is not proved obvious merely by demonstrating that each element was independently known in the prior art.

[00:12:34] Speaker 00:
What the board seems to be saying here is Smile existed and had some of it, and our patent discusses Java and that could do other things.

[00:12:42] Speaker 00:
And therefore, since these two things exist in the prior art, Java existed and Smile existed, let's just call it close enough.

[00:12:49] Speaker 00:
But what you actually have to get to is a reason to modify or combine to get to the claimed control information file.

[00:12:55] Speaker 00:
And if Smile's not being modified, what is?

[00:12:59] Speaker 00:
Where does that control information file exist?

[00:13:02] Speaker 00:
It's not clear either from the petition or from the board's final written decision where that exists.

[00:13:09] Speaker 00:
And absent a showing on that, and absent some reliance

[00:13:12] Speaker 03:
I think I'd like to begin, unless the court would like to direct me somewhere else,

[00:13:36] Speaker 03:
by trying to clarify here just what the board found, and this is a substantial evidence review case of course, and what the board found was taught in the smile reference itself versus what needed to be supplied.

[00:13:50] Speaker 03:
This whole discussion of controlled information files the board found expressly to have been taught in the smile reference.

[00:13:56] Speaker 03:
That's an appendix 15 and 16.

[00:14:00] Speaker 03:
You know, we ticked through how the board had made findings on steps one through five, method steps one through five, in the SMILE reference itself in our red brief.

[00:14:12] Speaker 03:
I had understood the dispute here to be about where there would be a teaching of step six, concurrent with the media presentation retrieving a next file.

[00:14:22] Speaker 03:
That is the one and only step in this claim that the board found not to have been taught in SMILE, the reference itself, and therefore look to other teachings in the art.

[00:14:33] Speaker 03:
The way it looked to other teachings in the art wasn't perfectly appropriate.

[00:14:36] Speaker 03:
It found, based on both the testimony of Dr. Balterman and the Hua reference, that there would have been a motivation.

[00:14:44] Speaker 03:
The concept here is very simple.

[00:14:45] Speaker 03:
It's simply while you're starting to play one portion of a media presentation, downloading the next present section.

[00:14:52] Speaker 03:
that concurrent with the media presentation retrieving the next file, Hua's talk was conventional and Hua's talk a motivation.

[00:15:01] Speaker 02:
Are you now attempting to defend the board's decision to, and I'll make it really clear to telegraph this as a hostile question, go off the rails and exercise its own discretion to come up with a combination that's not in the petition?

[00:15:17] Speaker 02:
What you're defending?

[00:15:19] Speaker 03:
I'm not, I mean, I would say that that was perfectly appropriate to do, but that is not the thrust of my question.

[00:15:25] Speaker 02:
Okay, so let me just be clear.

[00:15:27] Speaker 02:
So on page 485, which is the institution decision, when the board says we exercise our discretion, an institute and IPR on an additional ground, that would be the combination of smile and hoa, correct?

[00:15:40] Speaker 03:
That is the combination of smile and hoa.

[00:15:42] Speaker 02:
Okay, so that is not a combination that is one of the two that you argued for in the petition?

[00:15:47] Speaker 03:
Well, I would argue, Your Honor, that it is, in fact, the same arguments and the same evidence.

[00:15:51] Speaker 02:
Except it's not, because you didn't.

[00:15:53] Speaker 02:
You have petition number one, anticipation in light of smile, number two, obvious in light of smile.

[00:15:59] Speaker 02:
And the Board itself actually found this was an additional ground, didn't it?

[00:16:04] Speaker 03:
The Board was clarifying what it understood the petition to have argued.

[00:16:08] Speaker 02:
Because I see here, nonetheless, for clarity, we exercise our discretion and institute an IPR on the additional ground.

[00:16:17] Speaker 02:
You don't think?

[00:16:18] Speaker 02:
Then I should interpret that as the board deciding it was going to add its own ground to the petition.

[00:16:25] Speaker 03:
If your honor does.

[00:16:27] Speaker 02:
And I do.

[00:16:27] Speaker 03:
Okay.

[00:16:28] Speaker 02:
And I think it's really unreasonable for you to argue otherwise.

[00:16:32] Speaker 03:
And I'm not arguing that that is not denominated as a separate ground.

[00:16:36] Speaker 03:
The decision of the board on the first ground though was based on the same arguments and evidence and it's supported by substantial evidence.

[00:16:43] Speaker 02:
So are you defending the idea that the board has the discretion

[00:16:48] Speaker 02:
to institute an IPR on the ground of its own choosing?

[00:16:53] Speaker 03:
I am defending the notion that... It's a yes or no question.

[00:16:56] Speaker 02:
Are you arguing to me that the board has the discretion to institute an IPR on the ground of its own choosing?

[00:17:06] Speaker 03:
In this case on these facts, yes, but if Your Honor disagrees, I'd like to address how HWA relates to the first ground.

[00:17:13] Speaker 02:
You think yes.

[00:17:14] Speaker 02:
So you think SAS doesn't prevent you, doesn't prevent the director from exercising his discretion to come up with his own ground in the institution decision.

[00:17:25] Speaker 03:
I do think SAS goes to a different question, which is partial institutions.

[00:17:29] Speaker 02:
I'm not asking you whether SAS goes to a different question.

[00:17:32] Speaker 02:
I'm asking you whether SAS prevents it.

[00:17:35] Speaker 02:
Do you think the 17, and yes there are 17 because I counted them, statements in SAS,

[00:17:41] Speaker 02:
which indicate that the Director does not have exactly this sort of discretion.

[00:17:45] Speaker 02:
You think I should construe all of those as dicta, and that we the Federal Circuit are so well positioned vis-a-vis the Supreme Court that I should disregard all of them?

[00:17:53] Speaker 03:
I'm not asking you to disregard anything in sass, Your Honor.

[00:17:56] Speaker 03:
And if I may, let me explain why the decision below should be affirmed, even if Your Honor thinks that second ground is entirely ultra-various.

[00:18:02] Speaker 02:
Well, it's not whether I do.

[00:18:03] Speaker 02:
I want to understand what you're arguing about it, because I understood your brief to largely dodge it

[00:18:10] Speaker 02:
for I think the obvious reason that it was a pretty messed up thing for the director to do after SAS.

[00:18:19] Speaker 02:
But I do have a sense that your brief attempted to nonetheless defend it.

[00:18:24] Speaker 03:
Our brief did defend it because here what the board was doing was simply clarifying for Philip's own benefit that it wanted to hear about Hoa.

[00:18:30] Speaker 03:
But let me address the first ground because I do think that the facts here

[00:18:36] Speaker 01:
So I'll be clear, when you say first ground, there were two alternatives.

[00:18:39] Speaker 01:
Sorry, the first obviousness ground.

[00:18:40] Speaker 01:
The one that you put forward, which was obviousness, single, prior art, but in light of general knowledge, which included in your referenced walk.

[00:18:51] Speaker 01:
Correct.

[00:18:52] Speaker 03:
That's the ground to which I'm referring.

[00:18:54] Speaker 03:
Okay.

[00:18:54] Speaker 01:
And this... By the way, just to follow up on what Judge Moore was saying, you know, it's never been clear to me exactly when the board decides to intervene and when it doesn't intervene.

[00:19:06] Speaker 01:
This case stand should stand for the principle that the board thinks that it has discretion to come at a new ground in the petition.

[00:19:15] Speaker 01:
Did the board ever think about, you know, I mean, I don't know what I'm aware of.

[00:19:20] Speaker 03:
No effort to defend that.

[00:19:23] Speaker 03:
But I would submit here that it doesn't affect this court's disposition of the case for the simple reason that here the board found obviousness on

[00:19:33] Speaker 03:
the ground that I think everyone would agree is an appropriate ground which is smile in light of the general knowledge of the posa.

[00:19:41] Speaker 03:
The board at that very same passage that Judge Moore pointed to at Appendix 485 cited two cases for the proposition that it is perfectly appropriate to consider the general knowledge as reflected in other teachings of the art, the Randall case and the Ariosa case, and they both stand for the proposition that the board

[00:19:58] Speaker 03:
not only may but must consider references like wah that are not denominated as being part of a ground but are cited in the petition, are backed by evidence and argument in the petition and explain why a limitation of the claim, here the concurrent downloading and playback limitation, would have been met in the arc.

[00:20:18] Speaker 03:
The finding on that ground that is not the ground that is subject to Judge Moore's question is a finding that's supported by substantial evidence and demonstrates the obviousness of this claim.

[00:20:31] Speaker 02:
Well, so I guess my concern is I agree that there has to exist the ability to have single reference obviousness.

[00:20:38] Speaker 02:
Of course there has to exist it.

[00:20:42] Speaker 02:
So single reference obviousness necessarily means something is missing from the claim and that you have to look somewhere else to get it.

[00:20:48] Speaker 02:
So, of course, the place you should look is general knowledge.

[00:20:51] Speaker 02:
Would everyone know this is just an extension of what we've already been doing?

[00:20:56] Speaker 02:
But what this case feels more like to me is to reference obviousness, which I've made clear to you, I think, inappropriate because you didn't ask for an addition.

[00:21:05] Speaker 02:
But that's what I feel like it is.

[00:21:06] Speaker 02:
And I feel like somehow you're trying to avoid having to prove, A, motivation to combine wha and smile, because what you're doing is you're looking to wha for the missing element.

[00:21:18] Speaker 02:
But you're saying, but we're not really looking to WAAD for the missing element.

[00:21:22] Speaker 02:
We're just saying WAAD is an example of the fact that that missing element is a thing people generally skilled in the art would know about.

[00:21:30] Speaker 02:
And so I feel like you're backdooring.

[00:21:32] Speaker 02:
You're trying to get around having to prove a real obviousness case based on two references by saying, we're really going to look to the one reference to fill in really the general knowledge of a skilled artist.

[00:21:45] Speaker 02:
And we're not really using that reference.

[00:21:47] Speaker 03:
Your Honor, I would disagree with the characterization because I think what we're getting at here is that you don't need to just look at why.

[00:21:55] Speaker 03:
You could look at any number of references.

[00:21:56] Speaker 03:
The notion of downloading the next segment of a file while you're playing the file.

[00:22:02] Speaker 02:
But the PTO didn't say that, did it?

[00:22:03] Speaker 02:
The PTO holds that any number of references disclose this, and that's why it's

[00:22:07] Speaker 03:
It found that this was part of the general knowledge, and Hua itself stated that it was part of the general knowledge.

[00:22:13] Speaker 03:
This wasn't a new invention of Hua.

[00:22:15] Speaker 02:
Actually, Hua didn't say general knowledge.

[00:22:16] Speaker 02:
Don't misstate things.

[00:22:18] Speaker 02:
Be precise.

[00:22:19] Speaker 02:
Hua said conventional, but don't misstate things.

[00:22:22] Speaker 03:
Know your references.

[00:22:24] Speaker 03:
Apologies, Your Honor.

[00:22:25] Speaker 03:
It stated that it was part of, it was certainly characterized as a conventional technique in Hua.

[00:22:30] Speaker 03:
And Hua said, Hua provided a motivation, which we cited and the board cited,

[00:22:35] Speaker 02:
You were counsel below.

[00:22:37] Speaker 02:
Why didn't you, in your petition, argue smile and why?

[00:22:43] Speaker 03:
I don't know that I have an answer to why I didn't, other than to say, number one, we think it's anticipated for the reasons we articulated.

[00:22:51] Speaker 03:
We obviously did not prevail on that ground.

[00:22:53] Speaker 03:
But secondly, because under the Ariosa case and under the Randall case, I don't think it matters.

[00:23:01] Speaker 03:
Is, did we prove in our petition that the person of ordinary skill in the art reading SMILE implementing a SMILE player using the knowledge that the poser would bring to bear, would that poser be motivated?

[00:23:14] Speaker 01:
Well, you weren't trying to combine anything that's the invention or whatever, I mean, was, is this article, but you weren't trying to use the other stuff in law.

[00:23:24] Speaker 01:
Correct, Your Honor.

[00:23:25] Speaker 01:
So that would have been more of a combination if you were trying to use the other.

[00:23:30] Speaker 01:
I mean, WHA articles about all this new stuff they've come up with.

[00:23:33] Speaker 01:
That's exactly right.

[00:23:34] Speaker 01:
And their reference to the other stuff is just in terms of what the background.

[00:23:38] Speaker 01:
In fact, they criticize the inadequacy of conventional pipelining, which was kind of one of their points, right?

[00:23:45] Speaker 03:
Well, the WHA develops this two things.

[00:23:49] Speaker 02:
Could that possibly have been exactly the reason you didn't want to make the two reference obviousness?

[00:23:52] Speaker 02:
Because if you do, you've got a proven motivation to combine, and Hoa actually criticizes the very thing you want to extract out of it and say would have been used in this case.

[00:24:03] Speaker 03:
No, Your Honor, because Hoa makes clear that if you do this two-phase service model, that is what the bulk of Hoa is about, you can do fast forwarding, you can do rewinding.

[00:24:13] Speaker 03:
There's no requirement in the law that in order for a combination to be obvious, it should be advantageous for all the reasons every reference develops.

[00:24:21] Speaker 02:
Here's my problem.

[00:24:22] Speaker 02:
Our Arendy case,

[00:24:22] Speaker 02:
says that when you look to the general knowledge, you can do so when you're just talking about an extension of what is claimed in a single reference obviousness, but it should be much narrower when you're trying to fill in a missing limitation, which is what we have here.

[00:24:39] Speaker 02:
And so when you're trying to fill in a missing limitation, I find it troubling that you would look to a reference which in fact discourages pipelining to say they would use pipelining.

[00:24:51] Speaker 03:
If I may, Your Honor, let me address the Orendi point and then the discouragement point.

[00:24:55] Speaker 03:
On the Orendi point, Orendi is about the situation where an expert say so is unsubstantiated.

[00:25:01] Speaker 03:
That's what Orendi discusses.

[00:25:03] Speaker 03:
Here it is substantiated.

[00:25:04] Speaker 02:
It's substantiated by what?

[00:25:05] Speaker 02:
This is one of these things.

[00:25:07] Speaker 02:
You want to limit SAS to only a single sentence related to its express holding and not look at the 17 other sentences where the Supreme Court says the Director has no discretion.

[00:25:17] Speaker 02:
Likewise, you want to look at Arendt and say it's limited to this.

[00:25:20] Speaker 02:
Arendt actually very methodically goes through four separate things that you need to think about if you're going to go to the common knowledge, one of which is how significant the limitation is, all these different things.

[00:25:33] Speaker 03:
I would differentiate, Your Honor, between common sense, as in Orendi, and the general knowledge that the posa would have, that is substantiated by prior art here, Hua, which says this is part of the conventional knowledge.

[00:25:46] Speaker 03:
There's no issue of whether the prior art in fact teaches that downloading in parallel with playback would have been known.

[00:25:54] Speaker 02:
It says it's part of the conventional art, but it also says, but you shouldn't use it, and it goes through reasons to discourage my point.

[00:26:01] Speaker 03:
Let me address that point.

[00:26:02] Speaker 03:
On that point, the fact that prior art may be undesirable for certain purposes taught in your reference is not a reason why it doesn't... That doesn't simply make it a teach-away when those other reasons are found nowhere in the claim and have nothing to do with the claim.

[00:26:20] Speaker 03:
That's the Scorzi-Demitsu case, 870, 3rd, 1376.

[00:26:26] Speaker 03:
The Hua reference

[00:26:27] Speaker 03:
provides advantages of using the two-phase service model, which, by the way, one absolutely could do in a smile player.

[00:26:34] Speaker 03:
We just don't think you're obligated to do it in a smile player.

[00:26:37] Speaker 03:
And the point there is that the claims say nothing about the ability to fast-forward or rewind, which is the advantage that Wasmore complicated technique provides.

[00:26:47] Speaker 03:
All the claims require is a player, which the board found is taught expressly in Smile,

[00:26:55] Speaker 03:
That, when it is playing one segment, downloads the next segment concurrently.

[00:27:01] Speaker 03:
It's a very simple concept that Dr. Botherman testified and Hua states were within the conventional knowledge of someone of the art.

[00:27:08] Speaker 03:
The question is how you implement a smile player.

[00:27:10] Speaker 03:
The obvious way to implement a smile player, to avoid long wait time or highly jittering pictures, to quote from Hua, is to implement this conventional pipelining, this simple step six.

[00:27:23] Speaker 03:
We concede that that step six is not taught expressly in SMILE, but we think it's an appropriate single reference obviousness case because the obvious way to implement a SMILE player by a person of ordinary skill in the art informed by the general knowledge that this is just the way things are typically done and an advantageous way to do it would be to use that.

[00:27:44] Speaker 03:
The person of ordinary skill and the art doesn't need to go the next step to Judge Prost's point and use the two-phase service model, nor does the person of ordinary skill and the art need to use Hua's particular equations for optimizing how you divide up, and perhaps dynamically, perhaps statically, but optimally divide up the segments.

[00:28:04] Speaker 03:
All that the claims require, and therefore all that we need to establish was obvious, is that the person, when a smile player is playing,

[00:28:12] Speaker 03:
it would have been obvious to download the next segment while you're playing the first.

[00:28:16] Speaker 03:
And that not only was part of the general knowledge, it's something that our expert Dr. Baltimore had testified was something that was in fact out there in existing players.

[00:28:24] Speaker 03:
That's in Appendix 1418 and Appendix 1317 to 1318.

[00:28:28] Speaker 03:
So the board's finding that the person of ordinary skill in the art would have been motivated to do that sort of playback concurrently with retrieving a file, the finding of motivation

[00:28:41] Speaker 02:
What is the basis?

[00:28:42] Speaker 02:
Why would you be motivated to make that change?

[00:28:45] Speaker 03:
You'd be motivated to make that change to, and I resist the notion that it's a change, one would be motivated to do that in a smile player so that, to minimize wait times for the exact reasons that while taught, conventionally pipelining was used to avoid having, making users tolerate either a long wait time or highly jittering pictures.

[00:29:03] Speaker 03:
That's at Appendix 315 and Dr. Baltimore talked about it in Appendix 199.

[00:29:08] Speaker 03:
So we have here an expert witness pointing to a concrete prior art reference

[00:29:13] Speaker 03:
Explaining why the person of ordinary skill in the art would have been motivated to supply a missing feature, and then explaining that there would have been a reasonable expectation of success in doing so.

[00:29:25] Speaker 03:
There's a finding there too.

[00:29:26] Speaker 03:
Both of those findings are supported by substantial evidence, and I would submit to you that the situation is nothing like Arendy, where there was no such other reference.

[00:29:33] Speaker 03:
And the question was lack of substantiation.

[00:29:37] Speaker 03:
Here it's substantiated by Hwa, and that's appropriate under Randall and Ariosa.

[00:29:42] Speaker 01:
Thank you.

[00:29:50] Speaker 00:
Thank you, Your Honor.

[00:29:51] Speaker 00:
I'd like to touch on a few points.

[00:29:53] Speaker 00:
I think the issue here is that Google wants to have its cake and eat it too.

[00:29:58] Speaker 00:
They want to rely on Hua, but not have to explain how to combine Hua.

[00:30:03] Speaker 00:
And the reason for that, and a reason I think is important to realize, is that it's not only an issue of whether or not we're talking about the second part of Hua.

[00:30:13] Speaker 00:
The conventional pipelining in Hua is that complicated formula.

[00:30:18] Speaker 00:
My colleague tried to distinguish the two and say, oh, you're not going to use the complicated formula.

[00:30:22] Speaker 00:
But if you look in HWA at the description of conventional pipelining, it is that complicated formula and it is this dynamic operation.

[00:30:31] Speaker 01:
So is your view more that there was no motivation to combine the two or that the implementation was difficult if not impossible?

[00:30:39] Speaker 01:
Because that's what I heard you saying last time you were here.

[00:30:42] Speaker 00:
Our position is this, there is no explanation of how those two would be combined and they're incompatible for certain ways.

[00:30:50] Speaker 00:
First, staying with the Hua idea, Hua describes... Well, let's talk about pipelining.

[00:30:55] Speaker 01:
You keep talking about Hua, but the takeaway from Hua in terms of general knowledge was simply what people call conventional pipelining.

[00:31:03] Speaker 00:
The conventional pipelining is described there was that dynamic interaction.

[00:31:07] Speaker 00:
They are taking information on the fly.

[00:31:09] Speaker 00:
I forget the exact letters, but they're taking information about the current download rate, the current playback rate.

[00:31:15] Speaker 00:
The reason they're doing that is it's a custom communication between a server and a player, where the two are talking back and forth and say, what's your playback rate?

[00:31:25] Speaker 00:
What's your download rate?

[00:31:27] Speaker 00:
And then dividing and segmenting on the fly based on that.

[00:31:30] Speaker 00:
Now that's conventional pipeline, and that's what's described under the heading conventional pipeline.

[00:31:35] Speaker 00:
But the issue is this, and this is why I think it's been avoided.

[00:31:39] Speaker 00:
If you're going to use what is described in WHA as conventional pipeline, and it is this custom communication and custom segmentation based on this, why would you ever have a control information file?

[00:31:51] Speaker 00:
A control information file tells you in advance what your segments are and what your alternatives are.

[00:31:56] Speaker 00:
As was testified by Dr. Porter in

[00:31:58] Speaker 00:
In Hua's conventional pipeline there are no control information files because you don't know in advance what those segments are going to be.

[00:32:05] Speaker 00:
And there are no alternatives because you are creating custom segments.

[00:32:08] Speaker 00:
You don't have to have an alternative if you're actually tailoring it to this specific exchange.

[00:32:13] Speaker 00:
So the question becomes if someone's going to use what's in Hua, why would they ever use a control information file and why would they ever use alternatives?

[00:32:21] Speaker 00:
The ideas are incompatible and that's the problem and that's what's been avoided here.

[00:32:26] Speaker 00:
The argument that was offered here today is that, well, there are other things out there that teach pipeline.

[00:32:33] Speaker 00:
Okay, where are they in the record?

[00:32:35] Speaker 00:
The only reference in the record with respect to what conventional pipelining is, is that complicated operation in HWALT.

[00:32:43] Speaker 00:
And what the argument was below and in the appeal briefs here is that we're not going to use that.

[00:32:49] Speaker 00:
We're going to use something else.

[00:32:51] Speaker 00:
But if the only reference for conventional pipelining is the thing that they don't want to combine and admit in their briefing that they're not going to combine, specifically stating, there was no need therefore to show how Hua's specific approach for dividing media files into segments would be implemented in SMILE.

[00:33:09] Speaker 00:
We're basically saying we're not going to use the complicated definition of conventional pipelining in Hua.

[00:33:14] Speaker 00:
We're going to use something else.

[00:33:16] Speaker 00:
But we're not going to provide any other reference on that.

[00:33:19] Speaker 00:
So it's basically trying to say,

[00:33:20] Speaker 00:
We gave you a reference, but we're not using it, and we're not combining it.

[00:33:24] Speaker 00:
But we gave you a reference, so Orendi says it's OK.

[00:33:28] Speaker 00:
But if there isn't a way to combine there, then there's simply a failure to provide a basis for obviousness.

[00:33:33] Speaker 00:
And you can't do a pen run around.

[00:33:35] Speaker 00:
Thank you.

[00:33:36] Speaker 01:
Thank both sides.