[00:00:06] Speaker 03:
The first case for argument this morning is 172147, amazon.com versus Zitovault.

[00:00:13] Speaker 03:
Mr. Bagotel, good morning.

[00:00:23] Speaker 00:
May it please the court, Dan Bagotel, on behalf of the appellants, with me is my colleague Christina McCullough.

[00:00:28] Speaker 00:
This appeal turns on the construction of the claim term sessions and its application to the Feinberg reference.

[00:00:35] Speaker 00:
We all agree that a session involves communication between two entities, a client and a server, over some period of time.

[00:00:41] Speaker 00:
Where we disagree is whether a reference that discloses a session has to disclose protocol level details about how a session

[00:00:50] Speaker 00:
starts and ends.

[00:00:51] Speaker 03:
There's a little confusion in my mind because people were just using words and to a certain extent talking past each other.

[00:01:00] Speaker 03:
Do you agree that what's confusing to me is that the board used recognizable beginning or end and you seemed to say that's okay but then you're interpreting recognizable to be something other than how.

[00:01:14] Speaker 03:
And it's just a little confusing to me in terms of what exactly the board meant and whether they were incorrect.

[00:01:20] Speaker 00:
We did have a disagreement over claim scope.

[00:01:23] Speaker 00:
It is a claim construction issue.

[00:01:25] Speaker 00:
The board used the word recognizable, but there was really a dispute.

[00:01:28] Speaker 00:
I think it was clear even from the part that they quoted on page 13 of the appendix.

[00:01:33] Speaker 00:
We were fine with the board's original construction, but they said there was some ambiguity in the construction.

[00:01:39] Speaker 00:
And we agreed we're willing to go with what our expert says is that a session, by definition, is a communication over some period of time, which the board

[00:01:48] Speaker 02:
But I thought that you agreed to recognizable beginning and an end.

[00:01:52] Speaker 00:
Well, in the sense that there is a beginning and an end, the problem was the other side wanted to have protocol level details of exactly how it begins and ends.

[00:02:00] Speaker 00:
And we specifically said we did not agree to that.

[00:02:02] Speaker 04:
What about the board statement in its construction that it was not requiring a particular protocol or anything, it was just requiring a beginning and an end?

[00:02:12] Speaker 00:
Well, actually, I think what they were saying, they were not requiring a particular protocol such as IPSEC, but they were acquiring

[00:02:17] Speaker 00:
a protocol, that level of detail.

[00:02:21] Speaker 03:
And you say that because that's what they said was deficient in the piece of prior art, right?

[00:02:25] Speaker 03:
They refer to how.

[00:02:27] Speaker 00:
Oh, yes.

[00:02:27] Speaker 00:
At 22, at 23, at 26, they repeatedly required that level of detail.

[00:02:33] Speaker 02:
But the question is, did they require it, or I thought that they basically were saying, look, you haven't shown us a session that has a beginning and an end, much less how.

[00:02:43] Speaker 02:
how it would operate.

[00:02:45] Speaker 02:
But I didn't read that much less how it would operate as being built into the construction.

[00:02:51] Speaker 00:
Well, I think if you read Appendix 22, Appendix 23, and Appendix 26, you'll see that that's what they were saying was missing, that level of detail of precisely how it begins and ends.

[00:03:05] Speaker 00:
There wasn't a dispute that there were communications over time.

[00:03:08] Speaker 00:
We know that, for example, in Feinberg, there's

[00:03:12] Speaker 00:
a request for service and there are parallel requests for service because at some point the server gets busy and it redirects that to another server.

[00:03:21] Speaker 00:
So we know we have communications, we have a request for service and we have a response.

[00:03:25] Speaker 00:
And we have parallel sessions because it's been redirected to another server.

[00:03:29] Speaker 00:
So we know there's been communications over time.

[00:03:32] Speaker 00:
According to the board

[00:03:34] Speaker 00:
said we didn't show was precisely how that began and ended, the details of when.

[00:03:39] Speaker 00:
None of that matters to the claims because the claims are about horizontal scaling of sessions.

[00:03:44] Speaker 03:
That's what was confusing to me.

[00:03:46] Speaker 03:
Beyond it just being about scaling, it seemed like the problem the board was seeking to address, which I think is a fair problem, which is how do you distinguish between sessions?

[00:03:57] Speaker 03:
And it wasn't entirely clear to me, and maybe your friend can help us out on this, how

[00:04:02] Speaker 03:
its construction, whether it's recognizable then defined as how, addresses that.

[00:04:07] Speaker 00:
Well, I would agree that to the extent we have certain claims that require simultaneous or parallel sessions, you need to be able to say that there are multiple sessions.

[00:04:16] Speaker 00:
And you need to know for certain claims whether a session has been transferred to another server.

[00:04:21] Speaker 00:
But you don't need the level of protocol detail to know that.

[00:04:24] Speaker 00:
For example, we gave the example in our brief of a conversation.

[00:04:28] Speaker 00:
If somebody walks into the middle of a conversation

[00:04:31] Speaker 00:
They know that we're having a communication session without necessarily knowing how it began or how it will end.

[00:04:36] Speaker 00:
The same thing is true for this morning.

[00:04:38] Speaker 02:
But you need to know when it began and when it ended, right?

[00:04:42] Speaker 02:
Because otherwise, how do you know that there's a beginning and an end?

[00:04:45] Speaker 00:
Well, you know that there was a communication, some communication over time.

[00:04:48] Speaker 00:
For some patents and some claims, you may need that information.

[00:04:52] Speaker 00:
For these claims, it's simply about the fact of multiple sessions and the transfer of sessions.

[00:04:57] Speaker 00:
Example this morning, if somebody walks into the courtroom right now, they'll know that we're in an argument session.

[00:05:02] Speaker 00:
They won't necessarily need to know how it was gaveled to order and how you called me up to the lectern here.

[00:05:09] Speaker 00:
Although it's transferred when he comes up and the session is transferred to my friend on the other side.

[00:05:16] Speaker 04:
But where was that particular theory argued to the board?

[00:05:20] Speaker 00:
Well, we actually didn't propose a construction of session in our

[00:05:25] Speaker 04:
petition, but we explained... No, no, I mean an explanation of how it is that the prior Tata session and beginning and end almost inherently is what you seem to be saying.

[00:05:35] Speaker 00:
It is inherent because we actually explained... Where was that argued?

[00:05:39] Speaker 00:
I'm sorry.

[00:05:39] Speaker 04:
Where was that argued to the board?

[00:05:42] Speaker 00:
Well, in our petition and at the oral argument.

[00:05:44] Speaker 04:
You argued inherency in your petition?

[00:05:48] Speaker 00:
I don't know if we used that word.

[00:05:49] Speaker 00:
We certainly used that at the oral hearing.

[00:05:51] Speaker 00:
We're explaining that every communication

[00:05:53] Speaker 00:
If it involves a request for service and a response from the server to the client, that is going to be a communication over some period of time.

[00:06:02] Speaker 00:
Our expert used that.

[00:06:03] Speaker 00:
Dr. Rubin used that in his report.

[00:06:07] Speaker 04:
Do you have a particular site that you'd like to provide the court?

[00:06:11] Speaker 04:
Well, Dr. Rubin's definition of obsession was that- Not a definition, but application of the definition to the prior.

[00:06:27] Speaker 00:
Our petition did not focus on the definition of session, but where we applied it was in the petition.

[00:06:35] Speaker 02:
I thought your whole point with your due process argument was that this issue was never discussed below.

[00:06:43] Speaker 00:
No.

[00:06:44] Speaker 00:
No, I wouldn't say that because there was a dispute over their attempt to impose the word a defined beginning and end.

[00:06:52] Speaker 00:
And we weren't quite sure what that meant.

[00:06:55] Speaker 00:
what that level of detail of define was, there was quite a bit of discussion about that at the oral argument.

[00:07:02] Speaker 00:
And the board ultimately adopted the word a recognizable beginning and end.

[00:07:07] Speaker 00:
But the recognizable beginning and end.

[00:07:08] Speaker 02:
But you agreed to the recognizable part.

[00:07:10] Speaker 02:
You agreed to the recognizable part.

[00:07:11] Speaker 00:
We agreed to recognizable in the sense that you know that there is a beginning and an end.

[00:07:15] Speaker 00:
But we did not agree that there needs to be the protocol level of exactly

[00:07:20] Speaker 00:
how if there's some sort of a hello message or quietness.

[00:07:23] Speaker 02:
I mean, I looked at all those page sites that you gave us and there's nothing in there where they say you need protocol level of detail as to how it occurs.

[00:07:32] Speaker 02:
What they said is that you need to know that you can identify when it occurs.

[00:07:38] Speaker 00:
Well, when and how?

[00:07:39] Speaker 02:
I think if you look at Appendix 22... Well, that seemed to me to just be an extraneous statement.

[00:07:47] Speaker 02:
If you actually look at

[00:07:48] Speaker 02:
their repeated discussion of what the claim construction is and what Feinberg doesn't show, they go into detail about the fact that Feinberg just doesn't have a session that you can determine actually ends.

[00:08:07] Speaker 02:
They're saying if when the transfer occurs, that's when the session ends.

[00:08:11] Speaker 02:
There's nothing in Feinberg that shows a transfer.

[00:08:14] Speaker 00:
Well, let's take the transfer of sessions.

[00:08:16] Speaker 00:
Take a look at the patent at column 8, for example.

[00:08:19] Speaker 00:
That's where that's discussed.

[00:08:21] Speaker 00:
It's actually not even entirely clear in the patent.

[00:08:23] Speaker 00:
When you're redirecting a session in the patent, it basically says you're going to go from one server to another server.

[00:08:31] Speaker 00:
It doesn't even define exactly whether there's a brand new session.

[00:08:35] Speaker 00:
It talks about redirecting the session, but then it also talks about terminating the connection or the session with the first server.

[00:08:41] Speaker 02:
So it's an amalgam of the two.

[00:08:44] Speaker 02:
Recognizable beginning, recognizable end.

[00:08:46] Speaker 02:
When you agreed to that, what did you think recognizable meant?

[00:08:51] Speaker 00:
At the end of the day, you can determine that there was a communication over a period of time.

[00:08:57] Speaker 00:
We know that because there was a request and a response.

[00:08:59] Speaker 00:
So we know there was a communication over time.

[00:09:02] Speaker 00:
And we know that there was a communication over a period of time.

[00:09:05] Speaker 00:
I can argue and explain to you why we know it was actually using TCP.

[00:09:11] Speaker 00:
That requires a little bit more explanation through an expert.

[00:09:14] Speaker 00:
But we do know that there was a communication over a period of time.

[00:09:18] Speaker 00:
What Feinberg did not expressly disclose is what protocol Feinberg was using.

[00:09:22] Speaker 00:
Was there a hello message?

[00:09:24] Speaker 00:
Was there a goodbye message?

[00:09:25] Speaker 00:
Was it terminated just by a period of time?

[00:09:27] Speaker 00:
And our position is that's just not necessary for purposes of these claims.

[00:09:31] Speaker 02:
Well, are you able to recognize a beginning or an end in Feinberg?

[00:09:35] Speaker 00:
Yes, because we know that the communications or the session begins when there's a request for service.

[00:09:41] Speaker 00:
And then when the file is transferred, it's done.

[00:09:44] Speaker 00:
That's all they're doing.

[00:09:45] Speaker 00:
That's what Feinberg is about, is transferring code modules over the internet.

[00:09:52] Speaker 00:
There's been some disputes about whether it's using IP packets or Feinberg patents.

[00:09:57] Speaker 00:
That's really a red herring here.

[00:09:58] Speaker 00:
What matters here is that there is a communication.

[00:10:01] Speaker 00:
There is a session between a particular client and a particular server.

[00:10:05] Speaker 00:
It's over a period of time.

[00:10:06] Speaker 00:
Information is being transferred, and it's being

[00:10:10] Speaker 00:
rooted to a different server when a server is busy.

[00:10:13] Speaker 00:
That's exactly what the patent calls for.

[00:10:21] Speaker 00:
I'd like to add a couple of things.

[00:10:22] Speaker 00:
I mean, Xenovol has made the argument that you really do this need to know, you know, one endpoint needs to know what's beginning and what's ending.

[00:10:29] Speaker 00:
But if you look at the prosecution history, they eliminated the requirement of a defined protocol.

[00:10:35] Speaker 00:
They took it out.

[00:10:36] Speaker 00:
That's at Appendix 76566.

[00:10:38] Speaker 00:
They eliminated the requirement for a defined protocol.

[00:10:41] Speaker 00:
Of all the claims that remain, only one requires any sort of a particular way of beginning or end.

[00:10:46] Speaker 00:
That's claim nine.

[00:10:47] Speaker 00:
And that's not even an issue in this petition.

[00:10:49] Speaker 00:
So when they wanted to include a particular way of beginning or ending a session, they did so.

[00:10:56] Speaker 00:
And they didn't require it in the claims that are at issue here.

[00:10:59] Speaker 00:
We think that's enough.

[00:11:01] Speaker 00:
You should reverse the construction and remand for the board to reevaluate patentability under either the board's original construction or

[00:11:08] Speaker 00:
Dr. Rubin's construction, which is one or more communications exchanged between two entities over some period of time.

[00:11:16] Speaker 00:
May I reserve the rest?

[00:11:17] Speaker 03:
Good morning, Your Honors.

[00:11:32] Speaker 01:
Justin and I just received a call.

[00:11:35] Speaker 01:
I'd like to begin by addressing

[00:11:38] Speaker 01:
Something my friend said, actually in response to a question by Judge O'Malley asking, are you able to recognize a beginning and an end in Feinberg?

[00:11:49] Speaker 01:
And I want to focus on that question because in any type of communication, we live in a finite universe, there's always going to be a beginning and an end.

[00:11:57] Speaker 01:
There's got to be some physical beginning and end.

[00:12:00] Speaker 01:
But that's not the same thing as a recognizable beginning and an end.

[00:12:03] Speaker 03:
So can you point us when you're responding to Judge O'Malley's point to what in the intrinsic record, other than maybe the preferred embodiment, points us to a beginning and the specificity of a beginning or end?

[00:12:19] Speaker 03:
I'm sorry, can I back up a little bit?

[00:12:21] Speaker 03:
Do you agree that the board required not just, we don't know what recognizable means.

[00:12:28] Speaker 03:
Let's assume that there's a dispute over that.

[00:12:32] Speaker 03:
it is recognized?

[00:12:35] Speaker 03:
It didn't require a particular protocol, but did it require a protocol?

[00:12:40] Speaker 01:
No, no.

[00:12:40] Speaker 01:
What the board said, what it tried to do was find out what does a person of skill in the art understand the word session to mean.

[00:12:46] Speaker 01:
And what that word refers to is not just the communication in general, it refers to a specific type of communication.

[00:12:54] Speaker 01:
And so what they wanted to see is in Feinberg, is it using that particular type of communication?

[00:12:59] Speaker 01:
Is it using sessions?

[00:13:01] Speaker 01:
One way to explain whether or not you're using sessions would be to say, we should use SSL connections between the primary server and the secondary server, and then you would need to provide some additional technical detail about how you handle the SSL keys and that kind of thing.

[00:13:14] Speaker 01:
Another way would be to say that we're using some sort of proprietary mechanism.

[00:13:19] Speaker 03:
So you agree that the board required that the construction of recognizable is you have to articulate how it's done.

[00:13:29] Speaker 01:
The board did not require that a prior art reference has to explain how you recognize beginning and the end.

[00:13:34] Speaker 01:
What it said is that the reference has to be using the type of communication where the endpoints know the beginning and the end of the communication.

[00:13:42] Speaker 01:
And maybe I can provide an example.

[00:13:43] Speaker 03:
Well, how do you do that?

[00:13:46] Speaker 01:
Well, one way would be to be using an already known protocol.

[00:13:50] Speaker 01:
So using SSL or TCP, something like that, where everyone knows.

[00:13:53] Speaker 03:
So it has to articulate a known protocol in order to satisfy the construction of session?

[00:14:00] Speaker 01:
I think the difficulty I'm having answering your question is that I don't know that I can define all the hypothetical ways that a prior reference could meet this particular definition.

[00:14:11] Speaker 04:
Do you have any hypothetical ways that would not involve a protocol?

[00:14:16] Speaker 01:
Sure.

[00:14:16] Speaker 01:
So one of the examples we talked about the hearing below was using a UDP type communication.

[00:14:22] Speaker 01:
And let me back up and talk a little bit about the TCP example.

[00:14:25] Speaker 01:
We also discussed that below.

[00:14:26] Speaker 01:
And then I'll contrast that with UDP.

[00:14:28] Speaker 01:
So in a TCP or SSL, these are examples from the patent specification at, for example, column two.

[00:14:36] Speaker 01:
The way that works is when sender A wants to have a conversation or communicate with B,

[00:14:41] Speaker 01:
It will first send a message saying, hello, I'd like to start up this new connection with you and begin a conversation.

[00:14:46] Speaker 01:
They agree on the parameters of the conversation.

[00:14:49] Speaker 01:
And then they know that there's a beginning point.

[00:14:50] Speaker 01:
OK, we have opened up a connection.

[00:14:53] Speaker 01:
We're now going to have this transfer of information.

[00:14:55] Speaker 01:
When one side wants to end it, there's various ways you could do that.

[00:14:58] Speaker 01:
One way would be to agree, one side is silent for five minutes.

[00:15:02] Speaker 01:
We end it.

[00:15:02] Speaker 01:
Another way would be to send a terminate message.

[00:15:04] Speaker 01:
I'm sure there's other ways.

[00:15:05] Speaker 01:
So that is sort of the prototypical session.

[00:15:08] Speaker 01:
Everyone agrees that that's a session.

[00:15:10] Speaker 01:
That's a recognizable beginning and end.

[00:15:11] Speaker 01:
different way of communicating over the internet using UDP.

[00:15:15] Speaker 01:
What the sender does is take a message, divide it up into little packets, address each one to B, and send them off.

[00:15:22] Speaker 01:
Once those packets are sent, A doesn't know if any of them arrive at the destination, doesn't know what order they arrive in, and doesn't know if B starts receiving them from the very first packet or the 10th packet or the 100th packet.

[00:15:36] Speaker 01:
So there's not a clear beginning where both sides know, ah, this is

[00:15:41] Speaker 01:
This is where the conversation began.

[00:15:42] Speaker 01:
And on the flip side, when B starts receiving these packets, it knows it's receiving more and more information.

[00:15:49] Speaker 01:
But it doesn't know if it's going to continue receiving that information for five minutes, or an hour, or 30 seconds.

[00:15:54] Speaker 04:
So you're giving an example of something that, in your view, doesn't have a recognizable beginning and end?

[00:15:59] Speaker 04:
Just to make sure I'm following you, is that what you're saying?

[00:16:01] Speaker 01:
Yes, yes.

[00:16:02] Speaker 03:
Can you point us to anything in the intrinsic evidence that sort of explains this or compels that, other than the preferred embodiment in the patent itself?

[00:16:10] Speaker 03:
Where is the, I mean, you're saying that it has to have a protocol.

[00:16:15] Speaker 03:
It just doesn't have to identify what protocol it is.

[00:16:21] Speaker 01:
Right.

[00:16:21] Speaker 01:
I don't think you necessarily need to identify a particular protocol.

[00:16:24] Speaker 01:
But you need to identify what then?

[00:16:26] Speaker 01:
What do you need to identify?

[00:16:27] Speaker 01:
That's what I'm... You need to clarify that the type of communication you're using is one that uses sessions as opposed to some sort of sessionless way of communicating.

[00:16:37] Speaker 03:
So can you point us to what in the specification in your patent does?

[00:16:41] Speaker 03:
Sure.

[00:16:42] Speaker 01:
I'll give you a couple of examples.

[00:16:44] Speaker 01:
But I do want to say that the specification was written for a person of skill in the art that presumably is familiar with what a session is.

[00:16:51] Speaker 03:
Yeah, but your friend could make the same argument with respect to construing the prior art.

[00:16:55] Speaker 03:
So that's not going to carry the day here, right?

[00:16:58] Speaker 01:
Yes.

[00:16:58] Speaker 01:
OK.

[00:17:00] Speaker 01:
If you look at column 7, line 23.

[00:17:08] Speaker 01:
There's a statement there that says, the client connects to the main server and authenticates using one of server's known authentication methods, 420.

[00:17:16] Speaker 01:
The main server determines if it can accept a new session based on its current available processor bandwidth.

[00:17:21] Speaker 01:
And it goes on to say, if the main server can accept a new session based on available processor resources, then it agrees on a secret session to see with the clients and begins the session.

[00:17:32] Speaker 01:
What we can draw from this statement is that,

[00:17:37] Speaker 01:
The mere fact that the client sent a communication to the main server does not mean that that is when the session began.

[00:17:45] Speaker 01:
The word session here is referring to the actual substantive conversation between the devices.

[00:17:51] Speaker 03:
But what it says here is the main server determines if it can accept a new session based on its current available processor bandwidth.

[00:18:03] Speaker 03:
So is that the level of detail that's required?

[00:18:07] Speaker 03:
I'm looking for what level of detail in terms of how it's recognizable appears in the patent.

[00:18:12] Speaker 03:
So is that what it is?

[00:18:17] Speaker 01:
I'm not sure that I could say in all cases what level of detail is required.

[00:18:21] Speaker 01:
And part of the difficulty is that Feinberg is direct to solving a different type of technical problem.

[00:18:28] Speaker 03:
Well, the problem in this patent is scalability, right?

[00:18:32] Speaker 01:
Yes, and it's specifically talking about the IT environment where you're managing VPN connections to a corporate network.

[00:18:40] Speaker 01:
And when you're using a VPN, there's this negotiation period where if an employee is at home or in a hotel room, they click the Connect VPN button, and his device or her device connects to the corporate servers.

[00:18:52] Speaker 01:
They agree on what this VPN tunnel is going to look like.

[00:18:55] Speaker 01:
Then it's open.

[00:18:56] Speaker 01:
And now, when that user is on his computer, it looks as if he is physically

[00:19:00] Speaker 01:
connected to those corporate computers.

[00:19:03] Speaker 01:
And so the session in that case mimics what that physical connection is like.

[00:19:08] Speaker 01:
And that's really what the patent is directed towards, is managing those types of connections and being able to address what happens when you have not just one VPN connection, but hundreds or thousands or tens of thousands.

[00:19:19] Speaker 03:
And so you pointed us to line 23 to clarify.

[00:19:24] Speaker 03:
To a certain extent, maybe we're talking about enablement rather than prior art and obviousness because you're asking, you're embracing the board's definition and under that definition you would need to have enabled it under your intrinsic evidence.

[00:19:39] Speaker 03:
What's hard for me is that the board really didn't refer to anything in the patent that it was relying on for its definition of recognizable recession or whatever you want to call it.

[00:19:51] Speaker 03:
So I'm asking, I'm looking for you to point

[00:19:54] Speaker 03:
us to something.

[00:19:55] Speaker 03:
Now, you gave us that the main server determines if it can accept a new session based on its current available processor bandwidth.

[00:20:04] Speaker 03:
Anything more specific than that in terms of how the determination is made under this patent with respect to when a session begins and when it ends?

[00:20:12] Speaker 01:
There's other examples I could point to, but they get to the similar point, which is that the word session in this patent is never used as a synonym for just any type of communication over the internet.

[00:20:23] Speaker 01:
It is just used to refer to the actual substantive conversation that happens after you establish that pipeline or that VPN or that SSL session.

[00:20:33] Speaker 02:
You both agreed to use of the term recognizable beginning and recognizable end.

[00:20:39] Speaker 02:
And even the experts, Dr. Rubin, seemed to contemplate that.

[00:20:45] Speaker 02:
What did you understand recognizable to mean when you agreed to that construction?

[00:20:53] Speaker 01:
I think the best way to explain it was what we tried discussing at the hearing itself, which is that the purpose of a session is so that you can determine, when you have data coming in, is that a part of this conversation that I'm having?

[00:21:04] Speaker 01:
Or is it something else over here?

[00:21:07] Speaker 01:
And that's really what everyone's sort of agreed on at the hearing.

[00:21:10] Speaker 01:
So that gets back to the TCP versus UDP example.

[00:21:13] Speaker 01:
If a device is receiving UDP packets, it doesn't necessarily know what conversation they fit into or when that conversation is coming to an end.

[00:21:21] Speaker 01:
And you can contrast that.

[00:21:22] Speaker 01:
with TCP or SSL or VPN connection, where as you're receiving this information, you know exactly what conversation it relates to.

[00:21:30] Speaker 01:
And you could even have, you know, you can imagine a web browser with multiple tabs open.

[00:21:34] Speaker 01:
You could have multiple conversations going on between the same devices.

[00:21:38] Speaker 01:
But the session allows you to keep them distinct so you know what conversation you're focusing on.

[00:21:43] Speaker 04:
Do I understand you to be saying that the system itself is able to recognize which session is which?

[00:21:49] Speaker 04:
Is that what you mean by recognizable beginning and end?

[00:21:52] Speaker 03:
Yes.

[00:21:53] Speaker 03:
Well, and can you point us to something in the intrinsic evidence that describes that in the patent specification?

[00:22:01] Speaker 01:
I would also point to the claims, because the way this claim construction issue was teed up was Amazon took the position.

[00:22:08] Speaker 03:
OK, well, can you just answer that question first, and then you can give us the history or the background.

[00:22:12] Speaker 03:
But you said there's something in the specification, and then there's the language of the claim.

[00:22:17] Speaker 03:
So can you show us what you're talking about there?

[00:22:20] Speaker 01:
Sure.

[00:22:22] Speaker 01:
In claim one, for example, it talks about a system for conducting a plurality of cryptographic sessions.

[00:22:29] Speaker 01:
And there's also talk of transferring sessions in claims five and seven and determining if a server can accept a new session, supporting incremental sessions in claims one and six.

[00:22:38] Speaker 01:
These are the same types of examples that the board focused on.

[00:22:41] Speaker 01:
And the reason why is that Amazon's position is that the word session just referred to communication over the internet.

[00:22:49] Speaker 01:
And that's because in Feinberg, the only theory they had was that we can find something that looks like a client, something that looks like a main server, something that looks like an agent.

[00:22:58] Speaker 01:
And we know that there's got to be some type of transfer of information between those entities.

[00:23:01] Speaker 01:
And so therefore, if the word session just refers to communication, we're fine.

[00:23:05] Speaker 01:
We meet up.

[00:23:06] Speaker 03:
OK, well, we understand the broadness.

[00:23:08] Speaker 03:
But in petitioners' reply briefs, so they described at least one version of what their position was.

[00:23:15] Speaker 03:
I want to know if you agree or disagree.

[00:23:17] Speaker 03:
They say, of course, the session begins and ends at some point.

[00:23:20] Speaker 03:
That's axiomatic.

[00:23:21] Speaker 03:
But that is not the same as requiring disclosure of the details of an initiation and termination process.

[00:23:29] Speaker 03:
And then they go on to say, you're really arguing that the prior art reference must disclose protocol level details, details regarding how communication is started and stopped.

[00:23:40] Speaker 03:
Do you agree with that?

[00:23:40] Speaker 03:
Was that just the differentiation between the two positions taken?

[00:23:45] Speaker 03:
I don't agree with that, but... So you weren't requiring disclosure of the details of an initiation and termination process that wasn't in your view about what recognizable meant?

[00:23:58] Speaker 01:
I don't believe there's any requirement that Feinberg should have to copy and paste the SSO standard in order to... Well, that's not what I asked.

[00:24:06] Speaker 03:
I didn't ask whether you have to copy and paste the standard.

[00:24:08] Speaker 03:
I asked whether or not you have to require disclosure of the details of an initiation and termination process.

[00:24:18] Speaker 01:
What they would need to show is that Feinberg used this type of communication referred to as a session.

[00:24:24] Speaker 01:
I imagine there could be ways.

[00:24:28] Speaker 03:
Can you repeat that?

[00:24:28] Speaker 03:
I'm sorry, I missed.

[00:24:29] Speaker 03:
What did you say that they would have to disclose that Feinberg does what?

[00:24:32] Speaker 01:
What they need to prove is that Feinberg was actually using this particular type of communication referred to as a session.

[00:24:39] Speaker 01:
And the way to do it, I don't think they need to... And how do you do that?

[00:24:41] Speaker 03:
Okay, that's what I'm asking.

[00:24:42] Speaker 03:
So you're saying...

[00:24:44] Speaker 03:
I just want some words to define recognizable.

[00:24:47] Speaker 03:
So you're saying that your position wasn't that there has to be a disclosure of details of the initiation and termination process?

[00:24:56] Speaker 03:
That wasn't your position?

[00:24:58] Speaker 03:
No.

[00:24:58] Speaker 03:
So what was your position?

[00:25:00] Speaker 03:
How detailed?

[00:25:01] Speaker 03:
What is the disclosure that's required?

[00:25:04] Speaker 03:
Can you just articulate that?

[00:25:04] Speaker 01:
If there were other indications in Feinberg that the system in Feinberg knows all the information that you need to know to be using sessions,

[00:25:14] Speaker 01:
That would have been it.

[00:25:16] Speaker 01:
What is that?

[00:25:16] Speaker 03:
Knows all the information you need to know?

[00:25:19] Speaker 01:
Right.

[00:25:20] Speaker 01:
If there was enough disclosure that, for example, the primary server in Findberg can distinguish between a beginning and an end of a conversation.

[00:25:29] Speaker 01:
And I can give you an example of that.

[00:25:32] Speaker 01:
My friend referred to when a client requests a file from Findberg and receives a file.

[00:25:38] Speaker 01:
And they've gone back and forth on the briefing here.

[00:25:40] Speaker 01:
In their opening brief, they said,

[00:25:42] Speaker 01:
that one request response, that's a session, because there's beginning and end.

[00:25:45] Speaker 01:
But then in their reply brief, they said, well, you know what, actually, there could be some times where you need to get more payment information or request additional access to the file, something like that.

[00:25:54] Speaker 01:
Maybe you need more code files.

[00:25:55] Speaker 03:
So what's the disclosure in the intrinsic evidence that's required?

[00:25:59] Speaker 03:
I mean, I guess I'm having a difficult time understanding the nature of the specificity of the detail required in this disclosure regarding termination and initiation.

[00:26:10] Speaker 03:
Where is it in the spec?

[00:26:12] Speaker 03:
Is it what you've already shown us on line 23, or is there something else?

[00:26:17] Speaker 01:
The specification also refers to IPSEC SSL.

[00:26:19] Speaker 01:
There's no dispute that those protocols, everyone agrees, all the experts agree, they have recognizable beginnings and ends.

[00:26:25] Speaker 01:
So I don't think there's any dispute that the patent itself discloses that type of... In the preferred embodiment.

[00:26:32] Speaker 03:
I'm looking for, but you're saying that you're not limited to a particular protocol.

[00:26:36] Speaker 03:
So I'm looking for a description of the details of the disclosure broader than just the identification of a particular protocol.

[00:26:45] Speaker 01:
The other site I gave you prefers generally to sessions, not specifically to SSL or IPsec.

[00:26:51] Speaker 01:
So there could be other mechanisms for achieving the functionality you need to be able to maintain those tunnels.

[00:26:58] Speaker 03:
And so what are the other mechanisms?

[00:27:01] Speaker 03:
There could be others, but are there any boundaries or details or limitations to that?

[00:27:13] Speaker 01:
The limitations are the same ones that were referred to by both experts in that the system has to be able to recognize, beginning of a session, end of a session, is this communication

[00:27:24] Speaker 01:
in this session that I want to know about, or is it some other unrelated communication?

[00:27:28] Speaker 01:
That's the type of functionality that distinguishes a session from more general communication.

[00:27:33] Speaker 04:
Earlier you had said that it did not require particular protocols or use of protocols at all, is what I understand you to say.

[00:27:41] Speaker 04:
And I had asked you for an example.

[00:27:44] Speaker 04:
And the example you gave was actually for something that didn't need sessions.

[00:27:48] Speaker 04:
Can you give me an example for something that does

[00:27:53] Speaker 04:
satisfy sessions but doesn't use a protocol.

[00:27:59] Speaker 01:
I understand the word protocol to refer just generally to an algorithm or an agreed set of boundaries for communication.

[00:28:08] Speaker 01:
So you could write your very own protocol or your very own set of steps for having communication.

[00:28:17] Speaker 01:
And that could still satisfy session.

[00:28:20] Speaker 04:
Do I understand correctly that what you're saying is that you have to have a protocol, it just does not have to be a particular known protocol?

[00:28:27] Speaker 03:
Yes, yes.

[00:28:30] Speaker 03:
Okay.

[00:28:31] Speaker 03:
Okay.

[00:28:32] Speaker 03:
Thank you.

[00:28:41] Speaker 00:
I think the discussion now has clarified that their position as the endpoints must know is really at the rules level.

[00:28:48] Speaker 00:
You have to know the rules.

[00:28:50] Speaker 00:
That's protocol.

[00:28:51] Speaker 02:
As I understood your expert discussion, he differentiated between communication generally and sessions particularly.

[00:29:00] Speaker 02:
And he said sessions is a term that is known in the field, and that he said that sessions is the conversation between endpoints and that generally would be with some known protocol, and he listed several.

[00:29:16] Speaker 02:
Why isn't that enough for the board to say that Sessions has a defined term and a defined meaning and you don't show us anything in Feinberg other than general communication, which Dr. Rubin said was not a session?

[00:29:29] Speaker 00:
Communications is a broader term of which Sessions is a subset.

[00:29:33] Speaker 00:
Communications can be broadcast.

[00:29:35] Speaker 00:
A session involves a communication between certain endpoints over some period of time.

[00:29:39] Speaker 00:
Depending on the claims, you might need to know that level of detail exactly how it begins and ends.

[00:29:45] Speaker 00:
My point is that for purposes of these claims, you do need to distinguish between parallel sessions so that you have multiple sessions, and you do need to know that a session has been transferred from one server to another.

[00:29:57] Speaker 00:
But it's all about the horizontal scaling, this sort of thing about exactly when it ends, this sort of vertical concept is nowhere to be found in these claims.

[00:30:06] Speaker 03:
Because the board was searching for, I think it was, they were looking for a way to distinguish between the sections.

[00:30:12] Speaker 03:
That was their prompt session.

[00:30:14] Speaker 03:
That was their problem, right?

[00:30:14] Speaker 00:
Well, they got hung up on a gigabyte level transmission about whether that's one session or multiple sessions.

[00:30:21] Speaker 00:
And that's just simply not what these claims are talking about.

[00:30:24] Speaker 02:
I do want to refer, though, to the- So you can see that

[00:30:27] Speaker 02:
that you have to be able to distinguish.

[00:30:28] Speaker 02:
I mean, if you go through this whole patent that says over and over and over that the main server is determining whether or not it's gotten to the request, whether it's got to send it to someone else, so that it has to be able to distinguish.

[00:30:41] Speaker 02:
So when you say, I agree, it has to be able to distinguish, how do you define that?

[00:30:47] Speaker 00:
Well, it's a communication between certain endpoints over time.

[00:30:50] Speaker 00:
If you get a request from a client and the server says, I'm too busy, I need to shunt this over to server B,

[00:30:57] Speaker 00:
then that has been redirected because now you have a different set of endpoints.

[00:31:01] Speaker 00:
And you know that this communication is occurring over some period of time.

[00:31:04] Speaker 00:
I direct you to column seven.

[00:31:05] Speaker 00:
Column seven actually helps us.

[00:31:06] Speaker 00:
If you look at column seven, it talks about the main server.

[00:31:10] Speaker 00:
I'm on line 63 of column seven.

[00:31:13] Speaker 00:
It says the main server maintains a list of connections, bracket, sessions.

[00:31:18] Speaker 00:
It's just equating connections with sessions.

[00:31:20] Speaker 00:
I agree communications is a broader concept, but the connection sessions is really just showing that you've got

[00:31:27] Speaker 00:
different endpoints communicating.

[00:31:29] Speaker 00:
And my main point here is that for purposes of these claims, you're talking about moving sessions to other servers so that you can have horizontal scaling.

[00:31:40] Speaker 00:
You're not talking about this kind of importance of when a session within a packet is beginning and ending.

[00:31:47] Speaker 00:
That's just not relevant to these claims.

[00:31:49] Speaker 00:
In fact, both experts, I'm sorry.

[00:31:50] Speaker 04:
I just want to ask you, if session and having a recognizable beginning and end really simply just means that you have to be able to distinguish between the sessions, the system itself has to be able to do that, where did you argue that Feinberg teaches that?

[00:32:06] Speaker 00:
I was going to give you that citation because I never got a chance to answer your question before and I wasn't trying to duck it.

[00:32:12] Speaker 00:
The best I can give you in terms of our pleadings, quote unquote, our petition and our reply below,

[00:32:18] Speaker 00:
was Appendix 82 and 83, and Appendix 282 is our reply.

[00:32:23] Speaker 00:
That's where we applied to.

[00:32:26] Speaker 00:
But part of the problem, and I will admit this, is that at that stage, both at the petition stage and in the reply stage, we hadn't had a construction like this from the board.

[00:32:34] Speaker 00:
The board agreed with us that if we showed basically the transfer of information between the servers and the clients, we had established a session at the institution stage.

[00:32:45] Speaker 00:
So we didn't get into great levels of detail

[00:32:48] Speaker 00:
on how a session began and ended.

[00:32:50] Speaker 00:
Had we done so, we could have actually explained why Feinberg really necessarily uses TCB, because this UDP that he's been talking about would never work for Feinberg.

[00:32:59] Speaker 00:
You've got to be able to pull all the information back together, and that only works with TCP.

[00:33:04] Speaker 04:
But at this time, you did know that the patent owner was taking a different position on claim construction.

[00:33:10] Speaker 04:
Is that correct?

[00:33:10] Speaker 00:
Yeah, in the sense that they said it was defined beginning and end.

[00:33:13] Speaker 00:
But you see the frustration that,

[00:33:15] Speaker 00:
well, even in front of, even today, but also in front of the board, there was disagreement.

[00:33:21] Speaker 00:
The very portion in which the board was discussing with Mr. Nomenadis, they had a discussion saying, well, if all they mean is that there's communication over time, that's good enough.

[00:33:29] Speaker 00:
We need to have the endpoints know that they're having a beginning and end.

[00:33:34] Speaker 00:
He's demanding more than what our council agreed to, Mr. Kinzel.

[00:33:39] Speaker 00:
And so we never had a meeting of the minds on what recognizable or defined

[00:33:43] Speaker 00:
And we're still having this dispute today.

[00:33:45] Speaker 00:
It's a question of, I think you just got to go back to what is the patent need?

[00:33:49] Speaker 00:
What's patent driving at?

[00:33:50] Speaker 00:
What are the sessions here trying to do?

[00:33:52] Speaker 00:
And if you can distinguish between two horizontal sessions and one session being transferred to another, you've got a session.

[00:33:59] Speaker 00:
Whether it's timed out, whether the session ends by a timeout or the session ends by a goodbye message or whether it begins with the transfer of a packet or it begins with a hello message, none of that matters here.

[00:34:12] Speaker 00:
There is some teaching in the patent about different possible protocols, but again, they're not binding themselves to any particular protocol.

[00:34:20] Speaker 00:
I think that's all I have.

[00:34:22] Speaker 00:
Thank you.

[00:34:22] Speaker 03:
We thank both sides, and the case is suspended.