[00:00:00] Speaker 04:
First is number 241020, personal audio LLC versus Google LLC, Mr. Haney.

[00:00:09] Speaker 00:
Good morning, Your Honor, Mr. Haney.

[00:00:11] Speaker 04:
Okay.

[00:00:13] Speaker 00:
Handles.

[00:00:14] Speaker 00:
Sorry.

[00:00:15] Speaker 00:
Quite all right.

[00:00:18] Speaker 00:
May it please the Court, the District Court erred in granting Jamol because it, number one, overlooked substantial evidence that supported the verdict of infringement under the Court's original claim construction, and two, because the Court changed its claim construction after the jury's verdict.

[00:00:35] Speaker 00:
The claim construction of sequencing file required a file received by the player and stored by the processer be used

[00:00:43] Speaker 00:
by the processor to control playback of each song in the ordered sequence and to respond to control commands.

[00:00:50] Speaker 00:
The parties agree that the playlist file downloaded by Google Play Music is called list items and that the playlist refers to list items and not the queue.

[00:01:02] Speaker 00:
There are admissions from three Google witnesses, in addition to the testimony of Dr. Elmeroth, who is a personal audio expert, that list items, that downloaded file, is directly accessed in response to a GOAT control command or a PLAGUE control command to commence playback of the first song and dictate the sequence of playback thereafter.

[00:01:24] Speaker 01:
Which of the use cases?

[00:01:26] Speaker 01:
The parties talked about four use cases.

[00:01:30] Speaker 01:
Which one of you

[00:01:31] Speaker 01:
primarily relying on here for us today.

[00:01:34] Speaker 00:
So one and two.

[00:01:36] Speaker 00:
The play command and the go command are use cases one and two as argued in the, we use that terminology in the summary judgment briefing and in the motion and limine.

[00:01:48] Speaker 00:
And we believe use case three is also established, but primarily one and two.

[00:01:52] Speaker 00:
We didn't use that terminology primarily at trial because we thought it might be confusing for a jury.

[00:01:58] Speaker 00:
So it's really play and go.

[00:02:00] Speaker 00:
And then use case three is check playable song.

[00:02:03] Speaker 00:
So you're relying on one and two.

[00:02:04] Speaker 00:
One and two primarily, yes.

[00:02:06] Speaker 04:
It's the real issue here whether list items controls the play in real time or whether

[00:02:18] Speaker 04:
it does so indirectly by communicating the information to the queue, which then does the control.

[00:02:24] Speaker 04:
Is that what we're really talking about here?

[00:02:26] Speaker 00:
I think so.

[00:02:27] Speaker 00:
I think that's a way it can be characterized.

[00:02:28] Speaker 00:
And the court made very clear that it doesn't need to be in real time in its pre-trial, its pre-Birdick planning construction, because

[00:02:38] Speaker 00:
The court said that, first of all, Google requested an amendment, an addition to the claim construction on the eve of trial on June 7th, which asked for to impose a requirement that the downloaded file be referenced when each song is played.

[00:02:55] Speaker 00:
The court declined to give that.

[00:02:57] Speaker 00:
And the prosecution history certainly doesn't impose a repeated reference requirement that the list items be referenced each time a new song is played.

[00:03:08] Speaker 04:
I'm a little confused as to what we're arguing about.

[00:03:10] Speaker 04:
Are we arguing about what the district court's claim construction was, or are we arguing what it should have been?

[00:03:19] Speaker 00:
We are not arguing the court's claim construction.

[00:03:21] Speaker 00:
We disagreed below with the court's claim construction, but we're not appealing the claim construction.

[00:03:26] Speaker 00:
What we are appealing is that the court modified its claim construction after the verdict, and that was the basis for its grant of JMAW.

[00:03:36] Speaker 00:
And specifically, the court did one of two things.

[00:03:38] Speaker 00:
And it's not entirely clear which one.

[00:03:40] Speaker 00:
It either said that you can't use copies.

[00:03:42] Speaker 00:
So the court's ruling was essentially, in the JMAW, was that a different file called the Q is used to control playback.

[00:03:53] Speaker 00:
But the court's claim construction pre-verdict makes very clear that

[00:03:58] Speaker 00:
that it's OK if the cue is used as long as the originally downloaded file, which is list items, is used as, quote, part of control.

[00:04:07] Speaker 00:
He said that.

[00:04:08] Speaker 03:
What about the clarification really on the eve of trial, where Judge Connolly made it clear that you can't have two files.

[00:04:17] Speaker 03:
Everything has to be one file.

[00:04:19] Speaker 00:
uh... i don't believe he made that statement i believe he he he said that as long as that you must start with the original file in response to respond to a control command you're talking about what i would call the first clarification but then there was a second clarification pre-trial was there not?

[00:04:36] Speaker 00:
there was not uh... there in fact he denied a second request for clarification pre-trial which was essentially along the same lines uh... that google asked for a curative instruction

[00:04:47] Speaker 00:
and we addressed this in the blue brief at pages four and five, they asked for a curative instruction along the same lines and he declined to get it.

[00:04:54] Speaker 00:
So the claim construction that we got at the end of the claim construction proceedings was the claim construction that was presented to the jury at trial.

[00:05:02] Speaker 00:
And only after the verdict was there an imposition of these requirements, either that only the original file be used, which is contrary to the construction that he gave to the jury,

[00:05:15] Speaker 04:
or that there's this repeated reference requirement that the downloader... I think it's really a question of whether it has to be, as I thought you agreed a moment ago, whether it has to be real-time controlled by list items or whether it can be communication of information to the queue, which is then stored and becomes part of the queue control.

[00:05:35] Speaker 04:
And I'm not sure why the real-time requirement isn't an appropriate gloss on the claim construction.

[00:05:45] Speaker 00:
Well, for two reasons.

[00:05:47] Speaker 00:
One, it wasn't in his original claim construction under this court's precedent in the Weiland case.

[00:05:53] Speaker 00:
You can't clarify your claim construction after the case goes to the verdict.

[00:05:58] Speaker 03:
And specifically... You can clarify before it goes to trial, right?

[00:06:02] Speaker 00:
Absolutely.

[00:06:03] Speaker 03:
And there was no such... And then, wasn't there somewhere where the district court said a theory of infringement based on a showing that only a second file would be the control playback of each song in the order to which one can respond to control commands is not consistent with my claim construction?

[00:06:19] Speaker 03:
A copy of a file is not the file, but rather is a new and distinct file, i.e.

[00:06:27] Speaker 03:
a second file.

[00:06:29] Speaker 03:
841071.

[00:06:32] Speaker 00:
So the court was not saying that a copy can't be used.

[00:06:37] Speaker 03:
He was saying that if you're only using the copy... This is the clarification I was talking about.

[00:06:42] Speaker 00:
I agree.

[00:06:43] Speaker 03:
You told me there was no clarification.

[00:06:46] Speaker 00:
There was no clarification to indicate that you can't use copies.

[00:06:51] Speaker 00:
There was a clarification that if you only use the copy,

[00:06:55] Speaker 00:
then that doesn't meet his instruction.

[00:06:59] Speaker 00:
And that's not what the evidence showed.

[00:07:01] Speaker 00:
The evidence showed that the list items was used directly in response to a Go control command.

[00:07:07] Speaker 03:
That's for the first song, right?

[00:07:09] Speaker 03:
Not for each song.

[00:07:12] Speaker 03:
Can you just answer that question?

[00:07:14] Speaker 00:
Yes, access to a go, or execution of a go control command or play command commences playback of the first song in the sequence and every single song thereafter.

[00:07:25] Speaker 00:
Each song thereafter in the absence of a further command.

[00:07:30] Speaker 03:
So, did Dr. Almorov ever say that?

[00:07:34] Speaker 03:
He did, he said... He seemed to be saying the cue, I thought, is the one that controls playback and responds to...

[00:07:43] Speaker 00:
So he said, and there's a quote on page 18 of the blue brief where he said that list items, that playlist, that originally downloaded sequencing file is used to control playback.

[00:07:57] Speaker 00:
He did describe how the data flowed through.

[00:08:00] Speaker 03:
It was a flowing data collection of data theory that the playlist data from the list items table gets loaded into the queue list, and then from there it goes to the queue.

[00:08:18] Speaker 03:
Where it's stored.

[00:08:20] Speaker 03:
And also where the control of the playback and the responding to control commands occurs.

[00:08:28] Speaker 03:
That was his theory.

[00:08:29] Speaker 03:
I don't know what file you're talking about when it comes to the list items table.

[00:08:37] Speaker 00:
So as to Judge Steig's comment that you interjected, what is stored, what is downloaded and stored is list items.

[00:08:45] Speaker 00:
There's no dispute about that.

[00:08:47] Speaker 04:
So there's no real time control by list items.

[00:08:52] Speaker 00:
There is.

[00:08:52] Speaker 00:
In response to a go or play command, you go directly back to list items.

[00:08:58] Speaker 00:
Once it's after it's downloaded and stored, you go directly back to list items to commence playback, and playback continues in that order specified by list items.

[00:09:08] Speaker 04:
You're saying it's not using the stored information to do the controlling?

[00:09:13] Speaker 00:
It is.

[00:09:14] Speaker 00:
It is.

[00:09:15] Speaker 00:
It's using the stored information in the site.

[00:09:17] Speaker 04:
I'm still confused.

[00:09:18] Speaker 04:
I thought it was clear enough that List Items is not engaging in real-time control.

[00:09:26] Speaker 00:
There's evidence from three witnesses that it is.

[00:09:28] Speaker 00:
And I don't want to lose Judge Chen's question, too.

[00:09:31] Speaker 00:
So I'll try to answer both.

[00:09:35] Speaker 00:
What happens is that in response to a go or play command,

[00:09:41] Speaker 00:
The programmer, the processor, goes directly to ListItems, which is the downloaded and stored sequencing file.

[00:09:47] Speaker 04:
It starts Playback.

[00:09:50] Speaker 04:
You just said it's downloaded and stored.

[00:09:53] Speaker 04:
So it goes to the stored file, right?

[00:09:56] Speaker 04:
It does, which is ListItems.

[00:09:57] Speaker 04:
So it's not real-time controlled by ListItems itself.

[00:10:01] Speaker 04:
It's controlled by a stored copy of the ListItems.

[00:10:07] Speaker 00:
ListItems is the stored file.

[00:10:09] Speaker 00:
Not a copy.

[00:10:11] Speaker 00:
List items is the downloaded and stored file.

[00:10:14] Speaker 00:
So it's not a copy.

[00:10:15] Speaker 00:
The copying comes after.

[00:10:18] Speaker 03:
Why did Omarov say that the queue is the sequencing file?

[00:10:25] Speaker 00:
Dr. Amaroff described how the system worked and how the data flowed from the list items, which was the first sequencing file, and then it flowed to subsequent files.

[00:10:37] Speaker 03:
He basically said there were three first sequencing files, right?

[00:10:41] Speaker 00:
One could characterize his argument that way.

[00:10:43] Speaker 03:
Because he was really more focused on just the data per se, not so much on a file per se.

[00:10:55] Speaker 03:
on the text, not the book itself.

[00:10:58] Speaker 03:
And then the text was getting copied into a new book, Q-list items.

[00:11:02] Speaker 03:
And then the data and text got copied into a third book called Q. That's fair.

[00:11:09] Speaker 00:
But remember that the court construed file as a collection of data.

[00:11:13] Speaker 03:
So it wasn't a collection of data with a file name.

[00:11:16] Speaker 03:
Where is the file name here?

[00:11:18] Speaker 00:
It's list items.

[00:11:20] Speaker 03:
And so what he described, and his flowing data theory is not... But where is there evidence that the list items controls the playback of each song?

[00:11:30] Speaker 03:
It might be evidence that it can control the playback of the first song of a playlist, but where is the evidence in the record where we can tell, oh yes, this thing called list items controls

[00:11:43] Speaker 03:
the playback of each song.

[00:11:45] Speaker 03:
Forget about few list items.

[00:11:46] Speaker 03:
Forget about Q. It's all about list items table.

[00:11:49] Speaker 00:
Yeah.

[00:11:49] Speaker 00:
The evidence is quoted at pages 18 through 22 of the Blue Brief.

[00:11:52] Speaker 03:
And what that evidence establishes is that... Well, I mean, where in the JA?

[00:11:58] Speaker 03:
Where is the quotable quote?

[00:12:02] Speaker 03:
Or from, I don't know, a Google document?

[00:12:06] Speaker 03:
Or Dr. Almorov that says, yes, list items

[00:12:14] Speaker 03:
Assuming it's a file, it not only receives and stores the data, but it also controls the playback of each song and responds to the devotee.

[00:12:27] Speaker 03:
Not the cue list.

[00:12:29] Speaker 00:
Yeah, so with respect to, there's the testimony cited on pages 18 through 22, when you pin, and this is, so on 19, there's testimony of a Google Witness resume.

[00:12:42] Speaker 00:
When you hit that go command, does Google Play Music use the list items table to seed the queue?

[00:12:48] Speaker 00:
And the answer is yes, it does.

[00:12:51] Speaker 00:
So it's doing two things.

[00:12:53] Speaker 00:
It's starting playback, and it's seeding the queue.

[00:12:57] Speaker 03:
Seeding the queue means copying the data into the queue.

[00:13:00] Speaker 00:
It does.

[00:13:01] Speaker 00:
It copies the data into the queue.

[00:13:02] Speaker 00:
But it commences playback, and it keeps playing.

[00:13:05] Speaker 00:
in the absence of a further control command.

[00:13:07] Speaker 00:
So you never have to go to the queue to decide to play the next song.

[00:13:10] Speaker 00:
It keeps playing in the order specified by the playlist, by the list items table in response to a play or go command.

[00:13:19] Speaker 00:
There's no reason

[00:13:21] Speaker 00:
Logically, there's no evidence that you would need to go back to the queue after you start playback unless the user wanted to change the order later.

[00:13:32] Speaker 00:
So in response to go or play, you start with the very first song in list items and you keep on playing.

[00:13:39] Speaker 00:
So list items, controls, playback of each and every song in the playlist.

[00:13:44] Speaker 03:
These two patents, they expired many years ago, right?

[00:13:47] Speaker 00:
They did expire.

[00:13:48] Speaker 03:
Is there any other pending litigation with respect to these?

[00:13:52] Speaker 03:
Long, expired passion?

[00:13:53] Speaker 03:
There is not.

[00:13:55] Speaker 00:
OK.

[00:13:56] Speaker 04:
So is what we're talking about here sort of what the definition of list items is, whether it's the stored data that you say constitutes list items and that does the controlling through the queue, or is it some other location that's the original list item?

[00:14:18] Speaker 00:
There's no debate there.

[00:14:19] Speaker 00:
There's no debate that list items is the downloaded and stored file that contains the playlist.

[00:14:24] Speaker 00:
That's not disputed at all.

[00:14:26] Speaker 00:
The issue is whether that file, that list items file that is downloaded and stored, controls playback of each song in the sequence.

[00:14:35] Speaker 00:
There's testimony that it commences playback of the first song

[00:14:38] Speaker 00:
and that it dictates the sequence in all subsequent data structures and that the music keeps on playing in the order specified by that downloaded list items file in the absence of a control command.

[00:14:51] Speaker 00:
And again, the judge clarified pre-trial and didn't change the clarification that as long as list items, that downloaded file, is part of control, then that meets his claim construction.

[00:15:03] Speaker 00:
He said that

[00:15:07] Speaker 00:
I think personal audio is right.

[00:15:08] Speaker 00:
Having gone back and looked at the lexicography in question, I don't believe it's the case that the use of the file precludes the use of another file to execute the control function.

[00:15:19] Speaker 00:
All I said was that the file had to be used to control it, but that doesn't mean it can't be used with something else.

[00:15:24] Speaker 00:
But it must be used as part of the control.

[00:15:27] Speaker 00:
And we're going to quote that.

[00:15:28] Speaker 03:
So are you saying that the list items table and the queue are being used together?

[00:15:33] Speaker 00:
Yes.

[00:15:33] Speaker 00:
List items are used first?

[00:15:36] Speaker 00:
Always.

[00:15:36] Speaker 00:
in response to a player.

[00:15:38] Speaker 03:
I'm just trying to understand, because I didn't quite see that being said by Almoroth anywhere.

[00:15:45] Speaker 03:
I mean, I see on A47271, he says the queue becomes the sequencing file.

[00:15:51] Speaker 03:
The queue is now the name of that collection of data.

[00:15:55] Speaker 03:
And it's the queue that's used for the user control period.

[00:15:59] Speaker 03:
Stop.

[00:15:59] Speaker 03:
That's it.

[00:16:00] Speaker 00:
Yeah.

[00:16:01] Speaker 00:
On page 18, there's a quote, and I'm over my time here, but there's a quote.

[00:16:06] Speaker 00:
where he says that that sequencing file, when he's referring to the list items, if you can look at its appendix pages, 47242 to 243, when he says that sequencing file is used for user control.

[00:16:20] Speaker 03:
So then what he goes to do is says... I think he might be talking about Q, because at the bottom of A47244 he says the Q items, and then the Q that's actually used for user control.

[00:16:34] Speaker 03:
I think it says like he seems like he's pretty consistent when he says the queue that's actually used for user control and then that will be the sequencing file it's like whatever sequencing file is it's sort of this shape shifter that moves between

[00:16:50] Speaker 03:
list items table, queue list, queue, and it's the queue that it keeps saying over and over again that's actually used for user control.

[00:16:57] Speaker 00:
He says both.

[00:16:58] Speaker 00:
He says both.

[00:16:59] Speaker 00:
He says that list items is used for user control.

[00:17:01] Speaker 00:
He says the queue is used for user control.

[00:17:02] Speaker 03:
Where does he say that?

[00:17:03] Speaker 00:
Where does he say list items are used for user control?

[00:17:07] Speaker 00:
On page 47242, we've quoted

[00:17:11] Speaker 00:
language there on page 18 of the brief, and in that very same answer earlier, in the very same answer, he's referring to list items as that sequencing file.

[00:17:22] Speaker 03:
You say 847242?

[00:17:26] Speaker 00:
Correct.

[00:17:29] Speaker 00:
Where?

[00:17:32] Speaker 00:
I will get that in my response, or I can get it from right now.

[00:17:37] Speaker 00:
Go ahead and get it.

[00:17:38] Speaker 00:
Pardon?

[00:17:38] Speaker 00:
Go ahead and get it.

[00:17:59] Speaker 00:
Which volume are the appendix?

[00:18:01] Speaker 00:
This is 3.3.

[00:18:03] Speaker 00:
Thank you.

[00:18:31] Speaker 00:
So, on...

[00:18:55] Speaker 00:
And so on page 242, and he says, and so this is at line 16.

[00:18:59] Speaker 00:
46242.

[00:18:59] Speaker 00:
47242.

[00:19:00] Speaker 04:
47242.

[00:19:04] Speaker 00:
And he says, and so there's the list table after that.

[00:19:07] Speaker 00:
So the list contains information about the lists.

[00:19:11] Speaker 00:
And in that table, and he talks about the various playlists.

[00:19:13] Speaker 00:
It's essentially the playlist at line 22 and in the position in the playlist.

[00:19:19] Speaker 00:
And then he continues on 47243.

[00:19:26] Speaker 00:
It's sent over a network.

[00:19:27] Speaker 03:
I'm sorry, what line are you on now?

[00:19:30] Speaker 00:
on line three as part of that same answer of page 47243 of the appendix.

[00:19:38] Speaker 00:
And he says it's sent over a network and then it's stored and then we'll eventually get to the third piece is where that sequencing file is used for control and navigating through the playlist.

[00:19:48] Speaker 03:
What does he mean by get to the third piece?

[00:19:52] Speaker 03:
Wouldn't that be the queue?

[00:19:55] Speaker 00:
It's the third piece of the court's claim construction.

[00:19:57] Speaker 00:
I think he's saying it's downloaded as piece one, stored as piece two, and then we'll eventually get to

[00:20:04] Speaker 00:
to wear that sequencing file, meaning referring to list items.

[00:20:09] Speaker 03:
To be consistent with what he says on the following page, that when he's talking about the third piece, he's talking about the third piece of the trinity of the list items table, queue list, and queue, i.e.

[00:20:20] Speaker 03:
he's talking about the queue there, just as he's talking about the queue on the next page when it comes to user control.

[00:20:27] Speaker 00:
I read his testimony.

[00:20:28] Speaker 00:
It says, it's sent over a network, step one,

[00:20:31] Speaker 00:
stored as step two, and then we'll eventually get to the third piece.

[00:20:35] Speaker 00:
So that's how I understood it, and that's how the jury was entitled to understand it, with all inferences being drawn in favor of the verdict winner.

[00:20:44] Speaker 00:
And certainly, just to be, so I'm not, so I'm entirely clear.

[00:20:49] Speaker 00:
Multiple structures, Dattre Elmeroth did testify that multiple structures are used to control playback.

[00:20:56] Speaker 00:
But

[00:20:57] Speaker 00:
Google's own witnesses and Dr. Amaroff did not testify inconsistently with that.

[00:21:03] Speaker 00:
They all said that control starts with list items.

[00:21:08] Speaker 03:
But 847243 is the strongest piece of evidence to that effect, right?

[00:21:13] Speaker 03:
I know you keep saying it, and you've been keeping on saying it for 21 minutes, but when it actually comes down to the record,

[00:21:21] Speaker 03:
It's this statement right here.

[00:21:23] Speaker 00:
I would say that's not our strongest evidence.

[00:21:25] Speaker 00:
Our strongest evidence is the testimony from three Google witnesses that says list items, which is the downloaded and stored file, is directly accessed in response to a go or play command.

[00:21:35] Speaker 00:
The strongest evidence is from Google's own witnesses.

[00:21:39] Speaker 04:
OK, I think we're out of time.

[00:21:41] Speaker 04:
We'll give you two minutes for a bottle.

[00:21:43] Speaker 04:
Mr. Bagatell.

[00:21:58] Speaker 02:
I think Judge Chen, you have it exactly right.

[00:22:10] Speaker 02:
Appeal has been a bit of a game of whack-a-mole.

[00:22:12] Speaker 02:
They've changed their theories at various times throughout the game.

[00:22:16] Speaker 04:
Help me to understand what list items is.

[00:22:18] Speaker 04:
Is it the stored file?

[00:22:19] Speaker 04:
Is that correct?

[00:22:20] Speaker 02:
List items is what comes down from the network and contains a jumble of files and it's stored in a table called list items.

[00:22:28] Speaker 04:
Okay, so if the stored file performs the three functions,

[00:22:32] Speaker 04:
then there's infringement, right?

[00:22:34] Speaker 02:
Yes, if list items performs, actually, the multiple functions, actually there were four, but if the list items performs all those functions, then it would infringe the sequencing file limitation.

[00:22:45] Speaker 02:
The problem is that the only evidence at trial was that there was this flowing data theory, that a portion of the data in list items is extracted and put into another file called queue items, and a portion of the data in queue items is extracted and put into

[00:23:02] Speaker 02:
a working file, a working playlist called the queue.

[00:23:07] Speaker 02:
And everybody's testimony was that the queue is what controls playback.

[00:23:12] Speaker 04:
But is the queue controlling the playback using the data and information from list items?

[00:23:20] Speaker 02:
It is initially seeded, quote unquote.

[00:23:23] Speaker 02:
That's the word, populated.

[00:23:24] Speaker 02:
So the data come over.

[00:23:26] Speaker 02:
Initially, if you hit play or go, some data will make its way over to the queue.

[00:23:32] Speaker 02:
And so that doesn't satisfy the claim construction because the district court made very clear based on the lexicography during the re-exam history that the, you can't just copy information over.

[00:23:44] Speaker 02:
You have to use the actual downloaded file to control playback of all songs and respond to control commands.

[00:23:52] Speaker 02:
And they don't contest that construction.

[00:23:54] Speaker 02:
everybody's testimony.

[00:23:56] Speaker 02:
There were four times when personal audio itself said it was the cue that controlled the playback of all the subsequent songs.

[00:24:02] Speaker 02:
I'll list them for you.

[00:24:05] Speaker 02:
Appendix 47244, we've already discussed.

[00:24:10] Speaker 02:
When he came back at 47271, he had a slide that appears at 46506.

[00:24:19] Speaker 02:
And then Mr. Hanley said it again in closing at 48.019.

[00:24:23] Speaker 02:
Our witnesses said the same thing.

[00:24:24] Speaker 04:
But is the seeded data that gets into the Q file doing the control?

[00:24:31] Speaker 02:
No.

[00:24:32] Speaker 02:
No.

[00:24:33] Speaker 02:
The actual control, every time it looks to the Q. So it's true that in response to the initial play and go.

[00:24:38] Speaker 04:
But is the Q using the seeded data now?

[00:24:41] Speaker 02:
is that you using the seeded data from the list file is using the copy data to initially start the play and go but every time it's going to return to look at the queue and see what's going on.

[00:24:55] Speaker 02:
The example they used in their brief about a steering wheel was actually pretty helpful.

[00:25:02] Speaker 02:
I think that's a good example.

[00:25:03] Speaker 02:
You need to continuously consult your steering wheel when you get out of the parking lot this evening.

[00:25:09] Speaker 02:
Because if you turn the car right and then you stop looking at where the steering is, you'll keep doing doughnuts in the parking lot.

[00:25:14] Speaker 02:
I find those analogies... Okay, but here's the point.

[00:25:18] Speaker 02:
You have to constantly consult the sequencing file.

[00:25:23] Speaker 02:
Why do we know that?

[00:25:24] Speaker 02:
For two reasons.

[00:25:25] Speaker 02:
One, it's in the claim construction, because it says the same file has to be used to control playback of all songs.

[00:25:33] Speaker 02:
Where does that come from?

[00:25:34] Speaker 02:
I actually take a look at first appendix of page 8207.

[00:25:38] Speaker 02:
That is where the lexicography was.

[00:25:42] Speaker 01:
Is this the re-exam?

[00:25:44] Speaker 02:
It's the re-exam history.

[00:25:45] Speaker 02:
And that's where they distinguish between a mere playlist, just a bunch of songs, and a sequencing file that actually controls playback.

[00:25:54] Speaker 02:
And they say it's not just a playlist, but rather a file of data that the player references

[00:26:00] Speaker 02:
when deciding what audio pregnant to play in response to the presence or absence of a control command.

[00:26:05] Speaker 02:
You have to continually consult.

[00:26:07] Speaker 02:
That's how this works.

[00:26:08] Speaker 02:
The sequencing file is what controls the playback of every song.

[00:26:14] Speaker 02:
And that's just not the way it works in the Google Play music.

[00:26:18] Speaker 01:
So, Mr. Bagdad, how would, in your view, can you explain to me how the list items file would have to work in order for there to be infringement here?

[00:26:30] Speaker 02:
It would have to go back and look at list items when deciding should I play song number two or song number three or song number eight.

[00:26:39] Speaker 02:
Every time it has to control the playback of each song and respond to any control command you give.

[00:26:44] Speaker 02:
Those are the requirements of the claim construction which they no longer dispute.

[00:26:47] Speaker 02:
So there has to be a reference back to list items.

[00:26:50] Speaker 02:
And there just wasn't any evidence of that because it isn't true.

[00:26:54] Speaker 02:
The only thing that they tried to argue is that... What's being referenced back to was the queue.

[00:26:58] Speaker 01:
Sorry?

[00:26:59] Speaker 01:
What's being referenced back to, you're saying, is the queue.

[00:27:01] Speaker 01:
Exactly.

[00:27:02] Speaker 04:
And the queue wasn't downloaded.

[00:27:03] Speaker 04:
What they're saying is that the queue is doing this based on information that comes from list items, right?

[00:27:11] Speaker 02:
they're saying that the information originally came down in the list items and then part of it was copied into the queue that was sitting in the quota.

[00:27:17] Speaker 04:
And it's that information that originally came from list items that's doing the controlling.

[00:27:22] Speaker 02:
And that's exactly what the district court said before trial would not work.

[00:27:25] Speaker 04:
But is that an accurate statement that it is the information that originally came from the list items?

[00:27:30] Speaker 02:
No, and what our witnesses said was that you initially populate the queue.

[00:27:34] Speaker 02:
So let's say you have a playlist of Beatles' greatest hits and

[00:27:39] Speaker 02:
You download that playlist and then you hit play.

[00:27:43] Speaker 02:
So it'll then take the entire 28 songs, actually I guess it's 19, however many songs, say it's 28 songs on the Greatest Hits album, and it will put them in that order and it'll eventually come over into the queue.

[00:27:55] Speaker 02:
But let's say you don't like a particular song.

[00:27:57] Speaker 02:
You don't like George Harrison songs and you want to change the queue.

[00:28:00] Speaker 02:
So it can change.

[00:28:01] Speaker 02:
The queue is what actually is being used to drive

[00:28:05] Speaker 02:
the ship there.

[00:28:06] Speaker 02:
It's going to look to the queue every time and see.

[00:28:08] Speaker 02:
The playlist that's downloaded is still there.

[00:28:11] Speaker 02:
It's still in list items.

[00:28:12] Speaker 02:
It's never changed.

[00:28:14] Speaker 02:
It's your working playlist.

[00:28:16] Speaker 02:
And that's what it's consulting.

[00:28:17] Speaker 02:
And that's deliberate because they don't want to use, they don't want to make a permanent change to the downloaded playlist.

[00:28:24] Speaker 02:
You've got that forever.

[00:28:26] Speaker 02:
But what you're using to actually control playback is the queue.

[00:28:31] Speaker 02:
And that's the only evidence

[00:28:33] Speaker 02:
uh... of how it works they're only uh... they're they're work use case for the shipping theory violated the claim construction but their theory of use cases wanted to also violence the construction because the construction says the downloaded file has to be used not a copy of it you're saying is that once the information is transferred from uh... downloaded list items file into the queue it's no longer list items it's not it becomes a little bit it becomes actually it becomes

[00:29:03] Speaker 02:
There are something called queue items, which is another table that's stored.

[00:29:07] Speaker 02:
And then you have a working playlist in working memory.

[00:29:09] Speaker 02:
And that's the queue.

[00:29:11] Speaker 03:
Assuming these different data structures are files, list items, queue list, and queue, nevertheless, they're all different files.

[00:29:23] Speaker 02:
They're different files.

[00:29:24] Speaker 03:
And that's why they need to have the same files.

[00:29:26] Speaker 02:
you need to have the same problem the reason that they have a traveling data theory was a risk was that they wanted to say basically this sort of bunch of data is moving along and first this is the sequence of file and this is the sequence of file and this is the sequence of file and the construction that this report was clear on that all along the the clarification before trial was not a change it made very clear from is straight out of the language in the re-exam but the other side is saying this morning i can't tell if this is a new argument yet

[00:29:57] Speaker 03:
It's the list items that is the sequencing file.

[00:30:02] Speaker 03:
And yes, it's getting a little bit of help from an additional file called the Q file.

[00:30:08] Speaker 03:
But there's nothing wrong with having some assistance from a second file in order to execute these different functions being performed by the first file because

[00:30:23] Speaker 03:
the judge's first clarification pre-trial permitted that kind of scenario.

[00:30:29] Speaker 03:
So that's what I think... That's not a new argument.

[00:30:34] Speaker 02:
Why is that?

[00:30:37] Speaker 02:
If there were evidence that list items and the queue were both being consulted with respect to every song, that might enrich.

[00:30:47] Speaker 02:
But there wasn't any such evidence because the only evidence was that the information was transferred into the queue and then list items was not used anything, not used further.

[00:30:57] Speaker 03:
But the GOAT command.

[00:30:59] Speaker 03:
You're drawing from the list items, right?

[00:31:02] Speaker 02:
You're drawing originally from list items.

[00:31:04] Speaker 02:
It's called seeding the queue, or populating the queue.

[00:31:06] Speaker 03:
So I guess that would be their theory.

[00:31:10] Speaker 03:
Maybe the first part of the function of controlling the playback of each song is kickstarted off by the list items, but then its list items is assisted

[00:31:24] Speaker 03:
No, listener is not assisted.

[00:31:29] Speaker 02:
That's where the problem is, is list items isn't assisted.

[00:31:31] Speaker 02:
All the expert testimony was consistent.

[00:31:34] Speaker 02:
Once it's populated, list items is forgotten.

[00:31:37] Speaker 02:
It's all about the cue that's used for control.

[00:31:40] Speaker 02:
Again, four different times personal audio said that at trial.

[00:31:43] Speaker 02:
Our experts said that again.

[00:31:45] Speaker 02:
The jury wasn't entitled to make that up.

[00:31:47] Speaker 03:
They just told me that you're

[00:31:49] Speaker 03:
Google witnesses gave the story.

[00:31:52] Speaker 02:
So coming back to my original question, what

[00:32:09] Speaker 04:
What you're saying is that in order to satisfy this claim limitation, List Items has to do this in real time, that seeding the information into the queue is not sufficient.

[00:32:24] Speaker 04:
That's the heart of the matter, is it?

[00:32:25] Speaker 02:
If real time means yes, it needs to be consulted as you're playing song 2, you've got to check to see what the queue tells you.

[00:32:32] Speaker 02:
Yes.

[00:32:32] Speaker 02:
I think we're on the same page about real time.

[00:32:34] Speaker 02:
But there could be a debate over how quick, but it needs to be every song you've got to check with the cue and you're not going back to list items.

[00:32:44] Speaker 04:
It's as if... Even though the information that the cue is using originally came from list items?

[00:32:50] Speaker 02:
Sure.

[00:32:50] Speaker 02:
I mean, a lot of things come down in list items.

[00:32:53] Speaker 02:
All of your playlists come down in list items.

[00:32:55] Speaker 02:
So there's a big source of information out there.

[00:32:59] Speaker 02:
But what's actually used to control playback, according to all the witnesses,

[00:33:04] Speaker 02:
both the engineers and the experts of both sides, and actually Mr. Hanley's closing, what's actually used for control, quote unquote, is the queue.

[00:33:12] Speaker 01:
Mr. Becatel, maybe I'm asking a question based on technical impossibility, and if that's the case, I apologize.

[00:33:19] Speaker 01:
But what would be the case if you moved the whole of list items and just put it in the queue?

[00:33:25] Speaker 02:
It wouldn't work if list items has multiple playlists, because the queue is basically like an array.

[00:33:34] Speaker 02:
It's basically just a list of music file numbers and you wouldn't be able to take your whole playlist and dump it all into the queue.

[00:33:47] Speaker 02:
So that's why you hit a particular playlist and it moves the information.

[00:33:53] Speaker 02:
small fraction of what's in list items gets moved into queue items.

[00:33:57] Speaker 01:
So if list items has 50 playlists in it and you move playlist 22, that's what you're just talking about.

[00:34:09] Speaker 02:
Exactly.

[00:34:10] Speaker 02:
And then that's the playlist.

[00:34:13] Speaker 02:
And let's say you go to the gym and you decide to have a shorter workout today because you're sitting on the bench.

[00:34:18] Speaker 02:
You can shorten that playlist and only play songs one, three, and five, but you won't be changing your regular gym playlist that you want to use tomorrow.

[00:34:26] Speaker 02:
That's the whole beauty of the queue.

[00:34:28] Speaker 02:
You have a working playlist.

[00:34:31] Speaker 02:
And that is controlled.

[00:34:34] Speaker 02:
That is what is used to control playback and respond to different control commands.

[00:34:38] Speaker 04:
So is it fair to say that

[00:34:40] Speaker 04:
I understand the theory that there's no real-time control by list items because it's seeding the queue.

[00:34:49] Speaker 04:
Is there any issue as to whether the seeded information is performing the functions here through the queue?

[00:34:57] Speaker 02:
the seeded, well the seeded, it's just data, it's not a command.

[00:35:01] Speaker 02:
This is like C++ code or whatever.

[00:35:05] Speaker 02:
So you've got information and it refers to something called mplaylist.

[00:35:09] Speaker 02:
The code name for the queue is mplaylist.

[00:35:13] Speaker 02:
So you're going to look at mplaylist and it's going to say do you want to go to the next song and it's going to check what the next song is in the queue.

[00:35:22] Speaker 02:
So the data are just

[00:35:24] Speaker 02:
They're either stored in a table, like in queue items, or they're stored in a working memory, as in the case of the queue itself.

[00:35:33] Speaker 02:
But there's no control there.

[00:35:35] Speaker 02:
The control is all by the program, and the program uses the information in the queue, not the information in the signers.

[00:35:41] Speaker 02:
And there just wasn't any dispute about that at trial.

[00:35:44] Speaker 02:
Again, this really was a game of whack-a-mole where they keep coming up with different theories.

[00:35:49] Speaker 02:
I don't know if you want me to address use case three.

[00:35:51] Speaker 02:
He mentioned it briefly, but it's

[00:35:54] Speaker 02:
It's all made up by the lawyers after the fact.

[00:35:58] Speaker 02:
It's just absolutely false.

[00:36:00] Speaker 02:
There simply is no continuous consultation of list items.

[00:36:04] Speaker 02:
That's the short version of it.

[00:36:06] Speaker 02:
There was a reference at trial to something called get audio ID and list ID.

[00:36:11] Speaker 02:
That's a function.

[00:36:12] Speaker 02:
It was not a reference to a list item ID, much less the list item ID that came from list items.

[00:36:19] Speaker 02:
It's an invention of the lawyers after the fact.

[00:36:23] Speaker 02:
there just simply was no evidence of consulting the queue with respect to playback of every song on the playlist, and that's what the consultant said.

[00:36:31] Speaker 03:
When it comes to a user wanting to skip around, and maybe skipping a song, is that controlling playback of the songs, or is that responding to control commands?

[00:36:44] Speaker 02:
If you define control commands, well, I think

[00:36:49] Speaker 02:
It's usually a combination of both, because you could have a control command that basically shuffles, or you could have a control command that says go or play.

[00:36:58] Speaker 02:
In Google Play Music, there are ways that you can add things to your queue, subtract things from your queue.

[00:37:07] Speaker 02:
Not all of that is covered by the claims, of course.

[00:37:09] Speaker 02:
The claims refer to things like the continual playback, which I could get to if you want to hear the alternative ground for a firmance.

[00:37:18] Speaker 02:
thought not.

[00:37:21] Speaker 02:
But yes, you have to control playback of every song and respond to control commands, and those are sort of, they are complementary requirements.

[00:37:30] Speaker 03:
Right, I'm just trying to understand the distinction between the two.

[00:37:34] Speaker 03:
My example seems like it fits both of those.

[00:37:39] Speaker 02:
I think in some cases you will be because, for example, if you have a control command that says now you want to shuffle, then that will also result in the queue taking a different shape.

[00:37:52] Speaker 02:
It'll be 1, 8, 7, 6, 4, whatever, and then it'll play it 1, 2, 3.

[00:37:57] Speaker 02:
But it will always consult the queue in doing that.

[00:38:00] Speaker 02:
That's the key point for present purposes.

[00:38:05] Speaker 04:
OK.

[00:38:05] Speaker 04:
Thank you.

[00:38:17] Speaker 00:
Just a few points.

[00:38:18] Speaker 04:
He has two minutes.

[00:38:19] Speaker 00:
Just a few points, Your Honor.

[00:38:21] Speaker 00:
Your Honor mentioned the skip command.

[00:38:24] Speaker 00:
The court decided in its jay maul that the skip forward limitation, which is claimed for, of the 187 patent was infringed.

[00:38:34] Speaker 00:
And that Claim4, which has the skip command, requires scanning forward in the received sequencing file.

[00:38:42] Speaker 00:
So we have a conclusive determination that's not challenged by Google on appeal that the sequencing file claim construction is met with respect to not only go and play, he didn't find that, but there's a conclusive finding by him that it's met with respect to the skip command, which has to go to the received sequencing file.

[00:39:01] Speaker 00:
So we have at least one

[00:39:04] Speaker 00:
Control commands that we know accesses the downloaded and stored list items table.

[00:39:10] Speaker 00:
Point number two.

[00:39:12] Speaker 00:
Mr. Evans testified at trial, Appendix 47056.

[00:39:18] Speaker 00:
If you click on the play button, it will start playback with the first song in the playlist.

[00:39:23] Speaker 00:
The evidence is that the playlist is list items.

[00:39:26] Speaker 00:
The playlist is not the queue.

[00:39:29] Speaker 03:
Another piece of... The playlist is the playlist data, right?

[00:39:32] Speaker 00:
The playlist is the playlist in the queue.

[00:39:34] Speaker 00:
And Google's own witnesses testified that the playlist is the playlist in list items.

[00:39:41] Speaker 00:
And Google's own witnesses admitted that playlist refers to the playlist in list items.

[00:39:48] Speaker 00:
And that's cited on page 22 of the blue brief.

[00:39:51] Speaker 00:
It's Meyer Patel, the technical expert, and list items is the table where we store our playlists.

[00:40:00] Speaker 00:
The same witness, Mr. Evans, testified, is it fair to say that the downloaded playlist, again, list items, is used to control the order of the songs that are used to playback?

[00:40:13] Speaker 00:
He concludes, after some pairing, yes.

[00:40:16] Speaker 00:
So we have evidence that it not only starts playback, it sees the cue and determines the order of playback.

[00:40:24] Speaker 00:
And in the absence of a control command, it keeps on playing.

[00:40:27] Speaker 00:
So that gets us to the repeated reference requirement.

[00:40:30] Speaker 00:
and which is this, that it needs to, you can need to continually go out of time, so just a couple more sentences.

[00:40:37] Speaker 00:
Thank you.

[00:40:37] Speaker 00:
Yeah, the court rejected a repeated reference requirement.

[00:40:40] Speaker 00:
We addressed that in the Gray v.

[00:40:42] Speaker 00:
15-16 and the re-exam prosecution history is inconsistent

[00:40:48] Speaker 00:
with a repeated reference requirement because it has two files, one of which is downloaded, 307, one of which is only created on the player, 305, and even though the control commands only refer to the file created on the player, which is 351,

[00:41:07] Speaker 00:
That still infringes.

[00:41:10] Speaker 00:
And that's exactly the way Google Play Music works.

[00:41:13] Speaker 00:
Do you have one file that stands out?

[00:41:14] Speaker 00:
I think we're out of time.

[00:41:16] Speaker 04:
Thank you.

[00:41:16] Speaker 04:
Thank you.

[00:41:16] Speaker 04:
Thank both counsel.

[00:41:17] Speaker 04:
The case is submitted.