[00:00:00] Speaker 04:
We have three cases for argument this morning.

[00:00:06] Speaker 04:
Mr. Tholander, is that right?

[00:00:08] Speaker 04:
When you're ready.

[00:00:22] Speaker 02:
May it please the court.

[00:00:24] Speaker 02:
The most significant issue in this appeal is the summary judgment

[00:00:28] Speaker 02:
non-infringement granted with respect to Microsoft customer Dell's use of IIS ARR.

[00:00:34] Speaker 02:
Now, the case did proceed to trial on a different product, Microsoft's proprietary front door system.

[00:00:41] Speaker 02:
And while that judgment rests on legal errors, which we've raised here, it represents just a small part of the infringement that was alleged.

[00:00:47] Speaker 02:
And the court below disposed of the bulk of this case when it decided on some judgment, really just a single short paragraph, that the Dell system

[00:00:57] Speaker 02:
There was no indication that the Dell's use of IIS ARR satisfied two plain elements.

[00:01:04] Speaker 02:
And that was dynamic regenerating element and the concurrently processing element.

[00:01:09] Speaker 04:
Where in the expert's declaration does it discuss those two limitations?

[00:01:17] Speaker 02:
Dr. Jones, our expert, talked about the dynamic web generation at 8798.

[00:01:27] Speaker 02:
And he explained that using the Dell materials, the configuration materials, and the architecture diagrams which are in there, those pictures are also at page 20 and 21.

[00:01:41] Speaker 01:
Can you identify which sentence you're referring to at 8798 right now, just so that we know exactly what support you're relying on?

[00:01:52] Speaker 02:
Sure.

[00:02:07] Speaker 02:
Now, some of this is confidential, so I want to be careful.

[00:02:12] Speaker 02:
But basically, if you look to about the middle of that first paragraph there, once an incoming request is load balanced and sent to the page server, that page server will take data from the data tier process and generate dynamic response.

[00:02:32] Speaker 02:
Now, there's no dispute that you don't call to a data tier.

[00:02:36] Speaker 02:
for a static web page.

[00:02:38] Speaker 02:
You don't want to do that if you're populating a page with dynamic material.

[00:02:41] Speaker 04:
Sorry.

[00:02:42] Speaker 04:
What about the concurrent processing?

[00:02:44] Speaker 04:
The concurrent processing.

[00:02:45] Speaker 04:
That's going to be the real problem for you on the Dell stuff.

[00:02:48] Speaker 04:
OK.

[00:02:50] Speaker 02:
On that one, I would point the court to 8751.

[00:03:01] Speaker 02:
And there, Dr. Jones

[00:03:06] Speaker 02:
opines that ARR is designed to and will.

[00:03:10] Speaker 01:
Just to make sure we're following you exactly, which exact paragraph are you following?

[00:03:15] Speaker 02:
Paragraph 125.

[00:03:18] Speaker 02:
Then you can, those first two sentences, basically.

[00:03:24] Speaker 04:
But where does it say Dell does that?

[00:03:27] Speaker 04:
It sounds like it's capable.

[00:03:28] Speaker 04:
I mean, isn't the problem here is this stuff is all clearly capable of it.

[00:03:33] Speaker 04:
But you actually have to show that there

[00:03:35] Speaker 04:
doing it to have direct infringement.

[00:03:38] Speaker 04:
Sure.

[00:03:38] Speaker 02:
And I would point the court there to the case study that Microsoft did of Dell's system.

[00:03:45] Speaker 02:
And that's found, you can find that at 9089.

[00:03:50] Speaker 04:
So just before we move there.

[00:04:03] Speaker 04:
This thing you just cited, these sentences you just cited us on AD7151 compared to F125 just showed the capability.

[00:04:10] Speaker 04:
They don't actually show that Dell does that.

[00:04:12] Speaker 04:
They don't say anything about Dell's actually doing that.

[00:04:16] Speaker 02:
Well, I disagree with that because what he cites to source code and he says that when you use ARR, while the page server is processing, the web server will continue concurrently to receive

[00:04:33] Speaker 02:
process and return to web pages.

[00:04:36] Speaker 02:
Is this source code that cited Dell source code?

[00:04:40] Speaker 02:
No, it's ARR.

[00:04:42] Speaker 02:
Again, isn't that the problem?

[00:04:44] Speaker 04:
Sure, this is capable of doing it, but it's also capable of operating in different ways, right?

[00:04:50] Speaker 04:
You have to show Dell actually does this.

[00:04:53] Speaker 04:
Yes.

[00:04:54] Speaker 02:
And if we can look at the case study?

[00:05:00] Speaker 02:
Did you say 9051?

[00:05:02] Speaker 02:
9-0, it's 90-87.

[00:05:03] Speaker 02:
OK, sorry.

[00:05:05] Speaker 02:
It's a 90-89.

[00:05:13] Speaker 02:
Actually, the whole study is relevant, but if you look at the very last paragraph on 90-89, it's printed sideways, so it's a little bit hard to see.

[00:05:26] Speaker 02:
There's a quote there where they talk about they handled 1,800 concurrent connections.

[00:05:32] Speaker 02:
Right, so you're asking us to draw an inference.

[00:05:34] Speaker 01:
Did you just say 1800 concurrent connections?

[00:05:37] Speaker 01:
That's right.

[00:05:38] Speaker 01:
Where's the word concurrent?

[00:05:40] Speaker 02:
Well, 1800 connections.

[00:05:41] Speaker 02:
So he's talking about our Windows Server 2000 scaled when it handed versus the 1000 connections handled by the old server.

[00:05:50] Speaker 02:
And what he's talking about is that as their expert explained that the system comes standard with

[00:06:01] Speaker 02:
1,024 concurrent connections.

[00:06:04] Speaker 02:
Now, on a high level, if we can just step back for one minute, this whole system, which requires three computers, a web server and two page servers, only really makes sense if you are concurrently processing.

[00:06:19] Speaker 02:
If you're only using one computer at a time, you don't need two computers or three computers.

[00:06:24] Speaker 01:
So the whole point- Where is that test?

[00:06:28] Speaker 01:
Where is that explained by your expert?

[00:06:32] Speaker 02:
Well, I mean, that testimony is, I think, in Dr. Jones explains that particular websites must frequently process thousands of current requests for dynamic content.

[00:06:46] Speaker 02:
That's at 96, 95.

[00:06:50] Speaker 02:
There was the, again, I mean,

[00:06:56] Speaker 02:
Does he spell it out maybe as clearly?

[00:06:59] Speaker 02:
I think, honestly, he didn't think this was going to be a dispute because of what I've just explained.

[00:07:06] Speaker 02:
Dell processes 3.2 billion requests per quarter.

[00:07:11] Speaker 02:
That's also in this case study.

[00:07:14] Speaker 02:
You can call it an inference.

[00:07:17] Speaker 02:
But again, this is summary judgment.

[00:07:20] Speaker 02:
The only question is, have we raised

[00:07:22] Speaker 02:
have we raised a genuine material of fact with respect to whether Dell does, in fact, concurrently process?

[00:07:30] Speaker 02:
And you have to review this evidence in the light most favorable to parallel networks.

[00:07:33] Speaker 02:
In other words, when they're saying, our demands on our system are so great that we had to up the standard 1,000 connections, and the point is you're running them concurrently.

[00:07:47] Speaker 02:
That's why there's 1,000 connections, if you only needed one connection anyway.

[00:07:52] Speaker 02:
But they say, the demands on our system were so great that we had to increase the standard 1,000 connections to 1,800 connections on a single server.

[00:08:03] Speaker 02:
Viewing that evidence in the light favorable to us, I don't think there's any possible way you can come to the conclusion that there's no concurrent process.

[00:08:09] Speaker 04:
But there's nothing that says this.

[00:08:10] Speaker 04:
I mean, I don't understand why you couldn't have got your expert to actually opine and say, based upon all of the possibilities of this, the way the system would operate and stuff,

[00:08:21] Speaker 04:
Dell must need this concurrent limitation.

[00:08:24] Speaker 04:
But he didn't say that.

[00:08:26] Speaker 04:
And you're asking us, based upon what's essentially attorney argument, to make that inference.

[00:08:32] Speaker 02:
I don't think that's right, Your Honor.

[00:08:33] Speaker 02:
If you go back to 8751, he doesn't say that in some cases ARR will do this.

[00:08:37] Speaker 02:
He says that in use, ARR will concurrently process.

[00:08:41] Speaker 02:
Are you saying that in all uses of ARR, it will do this?

[00:08:47] Speaker 02:
I think that's a fair reading of that testimony.

[00:08:50] Speaker 02:
I mean, I guess it's maybe metaphysically possible that you would only ever send one request.

[00:08:56] Speaker 02:
But that's not what he testified.

[00:08:57] Speaker 02:
He testified that it will do this.

[00:09:00] Speaker 02:
That's the testimony.

[00:09:02] Speaker 01:
Did you say 90-51?

[00:09:03] Speaker 02:
Sorry, 87-51.

[00:09:06] Speaker 02:
And then when he gets to Dell, he makes the clear connection and says, Dell has told us that they use ARR in this way.

[00:09:14] Speaker 02:
Where is that?

[00:09:16] Speaker 04:
I mean, I don't understand why you're not pointing us to that if he actually said something specifically about Dell concurrently processing.

[00:09:23] Speaker 04:
Well.

[00:09:25] Speaker 04:
Right.

[00:09:25] Speaker 04:
You're not going to find it, right?

[00:09:27] Speaker 02:
Yeah.

[00:09:27] Speaker 02:
He doesn't say Dell concurrently processes.

[00:09:29] Speaker 02:
Not in those words.

[00:09:30] Speaker 02:
But he's told us that ARR concurrently processes.

[00:09:34] Speaker 02:
And then he tells us that Dell uses ARR.

[00:09:38] Speaker 02:
So again, I don't see how that could, if you're taking every inference in our favor, it seems to be very

[00:09:45] Speaker 02:
that very easily satisfies that threshold standard.

[00:09:48] Speaker 02:
And I would say that on appeal, Microsoft does not appear to dispute that Dell processes a number of requests at the same time.

[00:09:56] Speaker 02:
They basically say that.

[00:09:58] Speaker 02:
They argue that, well, maybe it's not concurrent processing because maybe the web server isn't processing at the same time.

[00:10:05] Speaker 02:
Or maybe the web server doesn't do any of the processing ever.

[00:10:09] Speaker 02:
Or maybe the page server doesn't do any of the processing ever in the Dell system.

[00:10:13] Speaker 02:
But we know that's not true because, again, the court found on summary judgment that there was no dispute that the accused web servers do processing.

[00:10:22] Speaker 02:
And the testimony about the page servers that we looked at earlier shows that page servers do the dynamic generation.

[00:10:34] Speaker 02:
So in our view, if you look at this evidence in the light of the summary judgment view,

[00:10:42] Speaker 02:
where we get the inferences, I don't think there's any way that that doesn't raise a question of material fact with respect to whether Dell does concurrent process.

[00:11:00] Speaker 02:
If I can turn to the survey for a minute.

[00:11:05] Speaker 02:
Now, the district court found the survey was both irrelevant and unreliable.

[00:11:10] Speaker 02:
We believe there's clear error on both of those fronts.

[00:11:13] Speaker 02:
So with respect to relevance, the survey was designed to find whether Microsoft customers use it in this general configuration of IAS with ARR and accused load balancing algorithm and across two or more page servers.

[00:11:28] Speaker 02:
Now again, the district court, similar, found that it wasn't relevant.

[00:11:32] Speaker 02:
And really, because the court, I mean,

[00:11:36] Speaker 02:
said different things about how many claims were sort of implicated by this.

[00:11:40] Speaker 02:
But the two things that the district court pulled out were, again, the dynamic generation and the concurrent processing.

[00:11:47] Speaker 02:
So there wasn't evidence of those things.

[00:11:49] Speaker 04:
And you didn't ask in the survey about those specific limitations.

[00:11:53] Speaker 02:
They did.

[00:11:54] Speaker 02:
Again, there was a threshold question about dynamic generation.

[00:12:00] Speaker 02:
So before you got to the point

[00:12:02] Speaker 02:
You know, the threshold question was how many number of requests?

[00:12:05] Speaker 02:
What about the concurrent processing?

[00:12:07] Speaker 02:
The concurrent processing, right, there wasn't a specific question about concurrent processing.

[00:12:12] Speaker 02:
But again, 40% of the respondents indicated that they have systems that generate over 1 million dynamic web pages per day.

[00:12:20] Speaker 02:
And they use this configuration with multiple computers.

[00:12:25] Speaker 02:
So you're into your rebuttal.

[00:12:27] Speaker 02:
You can keep going or you can save.

[00:12:28] Speaker 04:
I'll reserve my time.

[00:12:29] Speaker 04:
Thank you.

[00:12:39] Speaker 00:
I would like to address the questions the court had about concurrently processing.

[00:12:46] Speaker 00:
What counsel says is that the sheer sort of number of requests must mean that there has to be concurrent processing taking place.

[00:12:53] Speaker 00:
And we don't dispute that there might be concurrent processing taking place.

[00:12:57] Speaker 00:
What we do dispute, and what the court so found, is that the claim is very specific as to how that concurrent processing has to be taking place.

[00:13:07] Speaker 00:
What has to happen is,

[00:13:08] Speaker 00:
that while the page server is processing the request for the dynamic web page, the web server has to be concurrently processing other requests.

[00:13:20] Speaker 00:
And that is where there is literally zero evidence, either from Dr. Jones, parallel networks expert, or from any documents.

[00:13:29] Speaker 00:
And in fact, we proffer at page 28 of the red brief three systems that would use ARR,

[00:13:38] Speaker 00:
in this weighted round robin that they talk about, none of which would be concurrently processing the way the claim requires.

[00:13:46] Speaker 00:
One system might be where all web page requests, both dynamic and static, are being processed by the page server.

[00:13:54] Speaker 00:
Therefore, you don't have the concurrent processing going on the web server.

[00:13:58] Speaker 00:
Another possibility is that a system where all dynamic web page requests are being prosecuted, prosecuted, processed by the web server.

[00:14:07] Speaker 00:
Therefore, you don't have the concurrent processing as required by the claim.

[00:14:11] Speaker 00:
And the third possibility is where all dynamic web page requests are being processed by a page server and other requests are being processed by a different server, not the web server.

[00:14:23] Speaker 00:
Those are three systems right there where the use of ARR does not result in the concurrent processing as described by the claim.

[00:14:31] Speaker 01:
Do I understand with respect to the survey that what the district court judge was saying is

[00:14:35] Speaker 01:
It's not enough to say that you have ARR because of these different ways that it could have been implemented?

[00:14:44] Speaker 00:
Exactly, Your Honor.

[00:14:46] Speaker 00:
In fact, what the judge did was he walked through the three limitations of the claim.

[00:14:51] Speaker 00:
So the claim has one limitation is routing, that it's required you route a request for a dynamic web page to the page server.

[00:15:00] Speaker 00:
And it's a very lengthy, it requires all sorts of steps, the routing limitation does.

[00:15:04] Speaker 00:
The next limitation is processing the request at the page server, while there's concurrent processing going on of another request at the web server.

[00:15:12] Speaker 00:
And the third is that the page server then dynamically generates a web page.

[00:15:18] Speaker 00:
And so what district court judge, well, I'm sorry, appellate court judge Jordan sitting by designation of the district court judge in this case, went limitation by limitation and said,

[00:15:30] Speaker 00:
You never asked that question.

[00:15:32] Speaker 00:
The survey asked.

[00:15:33] Speaker 01:
But the survey doesn't have to go through each of those claim limitations, does it?

[00:15:37] Speaker 01:
It's only in the odd circumstances like here where you would have to do that.

[00:15:42] Speaker 01:
Is that correct?

[00:15:43] Speaker 00:
The survey doesn't have to prove.

[00:15:45] Speaker 00:
I mean, under Vitamix.

[00:15:47] Speaker 00:
Absolutely.

[00:15:47] Speaker 00:
And Judge Jordan addressed Vitamix, saying the survey doesn't have to prove infringement, but it somehow has to be linked to the claim.

[00:15:57] Speaker 00:
And I want to read his language

[00:16:00] Speaker 00:
verbatim because he stated it so well.

[00:16:03] Speaker 00:
So in his granting of the survey, what he did is, at appendix 37, he said, the asserted claims require that a web server be used to manage requests for dynamic web pages.

[00:16:16] Speaker 00:
That's step one.

[00:16:17] Speaker 00:
He said, yet the survey didn't ask respondents if they used the accused products to manage requests for dynamic web pages.

[00:16:23] Speaker 00:
So there's no link there at all.

[00:16:25] Speaker 00:
Not just that it didn't show infringement, they just didn't even ask if they were using the accused products for that purpose.

[00:16:31] Speaker 00:
The next is the claims require the web server to concurrently process multiple requests.

[00:16:36] Speaker 00:
Yet the survey did not ask the respondents if they used the accused product to concurrently process multiple requests.

[00:16:42] Speaker 00:
And lastly, the claims require the use of a page server to dynamically generate a web page.

[00:16:48] Speaker 00:
Yet the survey did not ask the respondents if they used the accused product to dynamically generate anything.

[00:16:55] Speaker 01:
it was inherent that those three limitations were satisfied by the use of ARR, just hypothetically, okay?

[00:17:02] Speaker 01:
Then you wouldn't have to ask those specific questions in a survey.

[00:17:07] Speaker 01:
Isn't that right?

[00:17:08] Speaker 00:
Probably.

[00:17:09] Speaker 00:
The problem is that I can't agree with the hypothetical about it being inherent.

[00:17:14] Speaker 01:
I understand.

[00:17:15] Speaker 01:
I said hypothetically.

[00:17:16] Speaker 01:
I understand that's not your position in this case.

[00:17:21] Speaker 00:
I'm sorry, I didn't mean to interrupt, Your Honor.

[00:17:23] Speaker 01:
I just want to know if I'm understanding the case thought correctly and understanding Judge Jordan's concern about that survey correctly.

[00:17:32] Speaker 01:
I think what it is is that it's not enough to say that this is ARR because there's so many different ways to implement ARR that it's not necessarily going to satisfy the claim limitations.

[00:17:44] Speaker 01:
So you actually have to ask about the claim limitations.

[00:17:47] Speaker 01:
I just want to know if that's your

[00:17:49] Speaker 00:
And that is what we're saying, Your Honor, exactly.

[00:17:52] Speaker 00:
That if it's a situation like in Vitamix, for example, or the case involving the stirrers, where it would necessarily infringe, then you don't have to do that.

[00:18:02] Speaker 00:
But here, and Judge Jordan pointed it out in Appendix 38, the emissions are even more significant when one considers that the accused products can be used for a variety of non-infringing purposes.

[00:18:12] Speaker 00:
So that's exactly what Your Honor is saying.

[00:18:14] Speaker 00:
If you don't have that situation, then maybe you don't have to ask those questions.

[00:18:19] Speaker 00:
But if you do have this situation, then you have to be specific.

[00:18:22] Speaker 00:
Not just that ARR is being used, but it's being used in an infringing fashion versus a non-infringing fashion.

[00:18:29] Speaker 00:
So did I answer your honest question?

[00:18:31] Speaker 00:
Yes.

[00:18:31] Speaker 00:
Thank you.

[00:18:32] Speaker 00:
Thank you.

[00:18:33] Speaker 00:
I'm not sure the court had any other questions of counsel.

[00:18:38] Speaker 00:
So there's just a couple of issues that were brought up in the reply brief that we haven't had an opportunity to address.

[00:18:46] Speaker 00:
One is that in the reply brief, it's argued that for the first time that Judge Jordan brought up the fact that there had to be only some processing going on at the web server

[00:19:05] Speaker 00:
and not a significant amount or a substantial amount.

[00:19:09] Speaker 00:
And it was argued in the reply brief that this is new to parallel networks.

[00:19:13] Speaker 00:
This is the first time this came up.

[00:19:15] Speaker 00:
When in actuality, this is exactly what parallel networks argued to the judge during the summary judgment and Daubert motions.

[00:19:26] Speaker 00:
So if we turn to appendix 12495, there was a discussion about intercepting.

[00:19:34] Speaker 00:
at 12495 at page 74, this is parallel networks arguing to the court, you're going to necessarily receive and process, I'm sorry, this is page 74, line 21, and it's at 12495, you're going to necessarily receive and process at some level

[00:19:54] Speaker 00:
the request before interception occurs.

[00:19:57] Speaker 00:
You have to do that in order to make a decision whether to intercept it or not.

[00:20:01] Speaker 00:
So this is parallel networks explaining to Judge Jordan.

[00:20:05] Speaker 00:
So you're going to have to do some processing.

[00:20:07] Speaker 00:
And then if we go to page 75 at that same appendix citation, line 10, parallel network says to the district court, the key to the patent when you read it

[00:20:20] Speaker 00:
is that there has to be an interception that occurs.

[00:20:24] Speaker 00:
And I think this is what the three construction goes to.

[00:20:27] Speaker 00:
There has to be an interception that occurs before the web server does something substantive to it.

[00:20:34] Speaker 00:
So where Judge Jordan got the idea that there could be some processing at the web server in order to decide whether it needs to be sent to the page server, in other words, whether this is a request for a dynamic web page, there has to be some processing occurring at the web server.

[00:20:50] Speaker 00:
Nothing substantive, no substantive processing can occur at the web server.

[00:20:55] Speaker 00:
And so it's no surprise that Judge Jordan then, at that same appendix citation on page 77, line five says, right, yes, I've got your argument.

[00:21:05] Speaker 00:
I've got your argument that there has some level of processing has to occur.

[00:21:10] Speaker 00:
It has to occur in order for it to be sent to the interceptor.

[00:21:14] Speaker 00:
So when Parallel Network says in their reply brief that the first time they heard of this

[00:21:20] Speaker 00:
different levels of processing was in Judge Jordan's post-trial denial of Jamal, it's just simply not correct.

[00:21:29] Speaker 00:
That is not what happened.

[00:21:30] Speaker 00:
They argued it to the district court during the summary judgment motion, hence why he got the idea of that, that some crossing is allowed at the web server, but not substantial or significant or, as they characterized it, substantive.

[00:21:44] Speaker 03:
Was there an agreed upon claim construction of intercepting

[00:21:48] Speaker 03:
prior to that summary judgment colloquy?

[00:21:50] Speaker 03:
What did it talk about?

[00:21:53] Speaker 00:
It simply talked about that the request had to be intercepted before it was processed.

[00:21:59] Speaker 03:
And so then there became the... So if you take that plainly, it has to be intercepted before any processing.

[00:22:06] Speaker 00:
That's what we argued, that summary judgment for no direct infringement by Microsoft, that was rejected by Judge Jordan, who said, well, there has to be some processing, because how do you figure out

[00:22:17] Speaker 00:
If the request is a request for a dynamic versus static web page, you have to figure that out in order to decide whether you're going to route it to the page server.

[00:22:24] Speaker 00:
So that's how that whole discussion came about.

[00:22:26] Speaker 00:
But it was an agreed upon construction.

[00:22:28] Speaker 00:
And then we followed that construction at trial, where our expert and our fact witnesses opined that there was a substantial amount of processing that went on at the web server, which I would, unless the court has any further questions, conclude my remarks with then.

[00:22:43] Speaker 00:
That's one of the reasons the jury ultimately found

[00:22:46] Speaker 00:
that Microsoft doesn't infringe.

[00:22:49] Speaker 00:
So you can use ARR the way Microsoft uses it, and there is no infringement.

[00:22:54] Speaker 00:
And we are on three grounds.

[00:22:56] Speaker 00:
Intercepting that, in fact, we do a substantial amount of processing at the web server before the request is sent.

[00:23:04] Speaker 00:
Releasing, we never release the web server to process other requests, which is just what we were talking about vis-a-vis Dell.

[00:23:11] Speaker 00:
And lastly, that it has to be a single request, and our system uses multiple requests.

[00:23:17] Speaker 00:
So even if somehow this court decided that there was some evidence that should have let the Dell question go to the jury and that there simply isn't, what would have happened is the same result.

[00:23:31] Speaker 00:
The jury would have found no infringement.

[00:23:33] Speaker 00:
Because if Dell is using our system the way we use our own system, we didn't directly infringe.

[00:23:39] Speaker 00:
So we certainly weren't inducing Dell to directly infringe, since we weren't directly infringing.

[00:23:46] Speaker 00:
And unless the court has any other questions, I would give three minutes and 50 seconds back to the court.

[00:23:53] Speaker 00:
Thank you.

[00:23:54] Speaker 00:
Thank you very much.

[00:24:07] Speaker 02:
Thank you.

[00:24:08] Speaker 02:
For one thing, the trial was not on ARR.

[00:24:11] Speaker 02:
So if there is a question coming back about issue preclusion, that hasn't been raised yet.

[00:24:15] Speaker 02:
And they would need to show that the systems are identical for all three claims.

[00:24:20] Speaker 02:
If I could say about your comment, Judge Hughes, about the intercepting limitation, which was that the plain language says that it would have to be before any processing.

[00:24:31] Speaker 02:
And I'll submit that there are two ways to read that.

[00:24:34] Speaker 02:
And one is what you just described, which was Microsoft's first approach.

[00:24:37] Speaker 02:
And the other is that phrase, before the request is processed, means before the processing, present tense moves to the past tense is processed.

[00:24:52] Speaker 02:
And that's really the point.

[00:24:53] Speaker 02:
You intercept it before it's processed, before the process is completed.

[00:24:59] Speaker 02:
So in other words, when we said the substantive, that was a way of saying,

[00:25:03] Speaker 02:
Well, yes, you intercept it before the page is generated, essentially.

[00:25:08] Speaker 02:
We're not being inconsistent.

[00:25:10] Speaker 02:
And there's no limitation in there about a significant amount of handling.

[00:25:14] Speaker 02:
And really, at trial, they didn't point to any kind of generation that was done by the web server, that was done by the page server.

[00:25:22] Speaker 04:
But isn't the limitation on handling all done by the first part of that, which is you divert the handling before processing?

[00:25:32] Speaker 02:
You divert the handling.

[00:25:34] Speaker 02:
Well, then the question is, what is handling?

[00:25:37] Speaker 02:
And handling is that process.

[00:25:38] Speaker 04:
To me, it sounds like you're trying to blur the lines between handling and processing.

[00:25:42] Speaker 02:
The court treated those as identical.

[00:25:45] Speaker 02:
And if you read the JMAW order, it uses actually the word processing instead of handling.

[00:25:51] Speaker 02:
There made no distinction between those two terms.

[00:25:54] Speaker 04:
Well, then if that's the case, then why isn't the agreed upon construction, you divert it before any handling or processing is done?

[00:26:02] Speaker 02:
You divert the handling.

[00:26:04] Speaker 02:
You're handling it.

[00:26:05] Speaker 02:
You divert that before the claim is processed.

[00:26:09] Speaker 02:
So as I explained, before the processing is finished, you divert it.

[00:26:13] Speaker 02:
In other words, the web server is not generating the page.

[00:26:16] Speaker 02:
So it's doing what it needs to do.

[00:26:17] Speaker 04:
So if handling is the same as the processing, if the web server can handle it, i.e.

[00:26:21] Speaker 04:
process it substantially, then why is that diverting?

[00:26:27] Speaker 04:
Because it is not generating the web page.

[00:26:29] Speaker 04:
It's diverting it before the web page is generated.

[00:26:32] Speaker 04:
But that's not what the claim construction says.

[00:26:33] Speaker 04:
It doesn't say anything about diverting the handling before the web page is generated.

[00:26:39] Speaker 02:
It says before it's processed.

[00:26:41] Speaker 02:
That's right.

[00:26:41] Speaker 02:
And we know because Judge Jordan said that cannot be read to mean no processing occurs at the web server.

[00:26:50] Speaker 02:
So it must be read to mean that it is processed as a past tense.

[00:26:54] Speaker 02:
In other words, before the claim is processed.

[00:26:58] Speaker 02:
Thank you.

[00:26:59] Speaker 02:
Thank you.