[00:00:07] Speaker 02:
We have six cases on the calendar this morning, two patent cases from the PTAB, one from a district court, an appeal from the Armed Services Board of Contract Appeals, a case from the Court of Federal Claims on Military Pay, and an employee case from the Merit Assistance Protection Board.

[00:00:32] Speaker 02:
The last two are being submitted on the briefs and will not be argued.

[00:00:39] Speaker 02:
First case is Microsoft versus Enfish, 2015, 1734 at AL.

[00:00:46] Speaker 02:
There's a lot of AL there.

[00:00:51] Speaker 02:
Several cases here.

[00:00:53] Speaker 02:
Mr. Campbell.

[00:00:59] Speaker 01:
Good morning.

[00:01:00] Speaker 01:
May it please the court, Chad Campbell here along with my colleague Theodore Wimson representing Microsoft.

[00:01:07] Speaker 01:
I propose to begin with claim 32 and the board's decision with respect to why it concluded Microsoft's evidence was not sufficient to demonstrate anticipation of that claim.

[00:01:21] Speaker 01:
I'm starting at the record appendix A27.

[00:01:25] Speaker 01:
In the first full paragraph on that page, the board indicates that they conclude that the evidence provided by Microsoft

[00:01:37] Speaker 01:
is not sufficient including primary key column information from CIS columns table is not persuasive because it does not explain how OID determination by text searching as recited in claim 32 would be conducted.

[00:01:55] Speaker 01:
Claim 32 does not recite text searching, it recites text entry.

[00:02:01] Speaker 01:
That was just a mistake of claims code by the board and it affected the way it analyzed the claim.

[00:02:09] Speaker 00:
The argument that is made as it relates to claim 32, at least the initial argument that Enfish makes, is that you actually never argued to the board in exactly the same way that you argued here.

[00:02:26] Speaker 00:
Did you ever actually mention user queries in your argument to the board?

[00:02:30] Speaker 01:
We had two arguments that we presented to the board.

[00:02:33] Speaker 01:
The first one was based on the index information in system columns table.

[00:02:39] Speaker 01:
or in the system tables table.

[00:02:42] Speaker 01:
We also had a separate item that we pointed the board to and that was the fact that the OIDs that were determined in connection with claim 31 are stored as primary key information in both the system columns table and the system tables table and the storage of that information as primary key information was sufficient to

[00:03:08] Speaker 01:
to be information in a column that's defined that would permit the OIDs to be determined and Again, which I don't think you're really answering that question because but the focus of all of your argument below had to do with searching right not entry Okay, it's true that we did talk a lot about searching.

[00:03:28] Speaker 01:
I'm not I'm not disputing that however the claim requirement doesn't require a

[00:03:35] Speaker 01:
searching, it requires text entry, and the searching we were talking about begins with text entry.

[00:03:41] Speaker 01:
So the question becomes, if you're going to interact with Chang et al., you've got to have text entry.

[00:03:48] Speaker 01:
Is there information that's defined to be presented and contained in a column that will enable the determination of an OID when you do enter text?

[00:03:58] Speaker 01:
And the answer there is yes.

[00:04:01] Speaker 01:
In the patent, there is a description

[00:04:05] Speaker 01:
of both searching type information that would be stored in a column, but also the OIDs themselves being stored in a column.

[00:04:14] Speaker 01:
And the patent teaches in various places that when you enter text to create a row or to add a column, that you would end up with the determination of the OID and the storage of that OID in the column that's meant for it.

[00:04:32] Speaker 01:
For example, if we look...

[00:04:34] Speaker 00:
But did you even talk about the CIS columns as it relates to claim 32?

[00:04:39] Speaker 01:
We did.

[00:04:40] Speaker 00:
Where?

[00:04:42] Speaker 00:
I couldn't find it.

[00:04:45] Speaker 00:
Yes, as to claim 31.

[00:04:49] Speaker 01:
You did.

[00:04:49] Speaker 01:
So if we begin at A220 in our petition, we pointed out at the bottom paragraph that the CIS

[00:05:04] Speaker 01:
CIS columns table is shown in Chang's figure three.

[00:05:07] Speaker 01:
And then we said that the CIS tables table is shown in Chang figure two as detailed below and taught by Chang.

[00:05:15] Speaker 01:
Each table, both of them, each of them separately, satisfies the elements of claims one, two, 31, and 32.

[00:05:22] Speaker 01:
We went on when we talked specifically about claim 32.

[00:05:32] Speaker 01:
to point out, for example, on A234 that Chang stores the column information in the CIS columns table, the definition of which is also stored in the PACT descriptor PD of the CIS tables table, Figure 2.

[00:05:48] Speaker 01:
So we did point to both the CIS columns table and the CIS tables table as alternative ways in which you could find columns that were defined to hold information

[00:06:02] Speaker 01:
that would permit the determination of OIDs from text entry.

[00:06:06] Speaker 02:
Mr. Campbell, maybe you want to move on to one of the other issues.

[00:06:09] Speaker 02:
You've raised a lot of issues, and you have a quite limited amount of time.

[00:06:14] Speaker 01:
Thank you, Your Honor.

[00:06:15] Speaker 01:
Let me, if I could, just briefly move to claims 55, 56, and 60.

[00:06:22] Speaker 01:
Claims.

[00:06:23] Speaker 01:
Is it the index of claims?

[00:06:25] Speaker 01:
Yes, they are.

[00:06:26] Speaker 03:
And they don't involve the self-referentiality property of the rescue.

[00:06:30] Speaker 03:
They do not.

[00:06:31] Speaker 01:
That is correct.

[00:06:31] Speaker 03:
So as I understand what the board said on those, it said all you said in your petition and all that Dr. Hosking said was these two references are in the same

[00:06:51] Speaker 03:
area, field, space, it doesn't matter what the word is.

[00:06:56] Speaker 03:
What you didn't do, I think the board said, was to say, if a skilled artisan was looking at visual basic, the skilled artisan would have some reason identifying we could do better, there's a problem here, to look elsewhere.

[00:07:14] Speaker 03:
And that crucial bridge between visual basic and Salton

[00:07:22] Speaker 03:
is what you just didn't have evidence of.

[00:07:27] Speaker 03:
Why was that unreasonable reading of the evidence?

[00:07:32] Speaker 01:
For the following reason.

[00:07:33] Speaker 01:
The bridge that we provided and talked about to the board can be seen if you look at how the board treated the discussion of Claim 54, which is the independent claim from which these others depend, followed by

[00:07:51] Speaker 01:
the identification for claim 55, for example, of the specific teachings that Sultan provided.

[00:07:58] Speaker 01:
If you look at page A, 1659 of Dr. Hosking's declaration.

[00:08:07] Speaker 01:
1659.

[00:08:08] Speaker 01:
Yes, 1659.

[00:08:11] Speaker 01:
At the very top, in paragraph 450, he points out that visual basic

[00:08:19] Speaker 01:
which is a software program that programmers are going to use, discloses an index record called an index object.

[00:08:28] Speaker 01:
Then in paragraph 451, he goes on to point out that it has the capability, Visual Basic has the capability of using the index record to locate records according to a keyword.

[00:08:42] Speaker 01:
Now, it is a programmer's tool.

[00:08:45] Speaker 01:
and therefore you have to put something into the program in order to use it, he points out that those details about what you would put into it can be found in a number of places, but Salton's one of them.

[00:09:01] Speaker 01:
And in Salton in 452, he points out that virtually every commercial system available

[00:09:09] Speaker 01:
is based on an inverted file design.

[00:09:11] Speaker 01:
An inverted file design is another name for a keyword index where you go through and you pick out important words and then create records where you line up all the places where you can find that word.

[00:09:22] Speaker 01:
So Salton teaches how to do that, a very simple keyword index, and then he goes on and points out that if you had a program like Visual Basic where you have to put something into the index to make it work, it would be natural to go look

[00:09:38] Speaker 01:
at a textbook like Sultan that taught details about what you would put in there.

[00:09:48] Speaker 01:
And so the two together, the fact that you've got a program that in order to use has to be filled with something, very specific teachings in Sultan about what to do, and the fact that you have a known system

[00:10:06] Speaker 01:
being used exactly as it has been used widely, this idea of keyword indexing, is the bridge that gives us the common sense hook to put those two together.

[00:10:17] Speaker 00:
And where in the record did you make this precise argument to the board?

[00:10:21] Speaker 01:
So in A1659, that's Dr. Hoskins'

[00:10:30] Speaker 01:
That's Dr. Hoskins' report.

[00:10:32] Speaker 00:
OK.

[00:10:32] Speaker 00:
I understand that it's in the doctor's report.

[00:10:35] Speaker 00:
But the only place I could find your even discussion of this motivation to combine was it was in your petition, not in any of your briefs.

[00:10:43] Speaker 00:
And in there, it was very conclusory.

[00:10:46] Speaker 01:
So we would submit that, in fairness, you need to read both what we said about the independent claim 54

[00:10:56] Speaker 01:
and Claim 55.

[00:10:59] Speaker 01:
When we get to Claim 55 expressly, after we've already explained the indexing capability of Visual Basic in Claim 54, we point out at Appendix page A318 that Visual Basic includes this idea of being able to search for keys and then we point specifically to the salt and teaching about index structures

[00:11:24] Speaker 01:
We point to the figure that Dr. Hosking was discussing, figure 1-10, which teaches how to use a keyword index.

[00:11:33] Speaker 01:
And then we point out that you've got a program, a teaching of what to put in the program.

[00:11:38] Speaker 01:
The two go together naturally.

[00:11:41] Speaker 00:
What is your response to M. Fish's argument that Salton taught away?

[00:11:46] Speaker 00:
I know that you didn't respond to that below.

[00:11:48] Speaker 00:
What's your response now?

[00:11:50] Speaker 01:
Salton taught many things.

[00:11:51] Speaker 01:
It was a textbook.

[00:11:54] Speaker 01:
It taught full-text indexing.

[00:11:58] Speaker 01:
In other words, every word.

[00:11:59] Speaker 01:
And it also taught more limited keyword indexing.

[00:12:03] Speaker 01:
And we relied on the more limited keyword indexing, not the bigger full-text indexing.

[00:12:10] Speaker 01:
The patent at column 13 recognizes that those two things are different.

[00:12:16] Speaker 01:
Well, claim 55 requires full-text indexing, doesn't it?

[00:12:19] Speaker 01:
It requires keywords.

[00:12:22] Speaker 01:
Keyword indexing.

[00:12:23] Speaker 01:
If we look at column 13 of the patent, it very clearly goes through and talks first about full text extraction at line 45 of column 13 and then points out that there are other types of extraction, both automatic and manual.

[00:12:42] Speaker 01:
And so the idea of more limited keyword extraction

[00:12:46] Speaker 01:
being different from full-text extraction is taught in the 604 patent, and it's also taught in Sultan.

[00:12:52] Speaker 01:
We were relying upon the more limited one.

[00:12:54] Speaker 01:
So there isn't a teaching away.

[00:12:57] Speaker 01:
I see that I'm into my rebuttal.

[00:13:00] Speaker 02:
You may stay before you.

[00:13:02] Speaker 02:
Thank you.

[00:13:03] Speaker 02:
Mr. Rahman.

[00:13:08] Speaker 04:
Good morning, and may it please the Court.

[00:13:11] Speaker 04:
I intend to speak briefly on the issues raised in Microsoft's appeal.

[00:13:15] Speaker 04:
and spend most of my time on Enfish's cross appeal concerning the construction of object identification number, or OID.

[00:13:25] Speaker 04:
There's an overriding issue that wasn't addressed by Microsoft in its argument, and that's the Enfish 1 claim construction.

[00:13:34] Speaker 04:
In case 2015-1244, which I'll refer to as Enfish 1, this court did claim construction.

[00:13:40] Speaker 04:
That claim construction's binding on this panel

[00:13:44] Speaker 04:
according to this court's precedents.

[00:13:47] Speaker 04:
And the court construed the term a logical table.

[00:13:50] Speaker 04:
A logical table is a phrase that appears in each of the claims.

[00:13:55] Speaker 03:
Can I just double check?

[00:13:57] Speaker 03:
That was a case coming from district court?

[00:13:58] Speaker 04:
Yes, Your Honor.

[00:13:59] Speaker 03:
So Phillips could govern that one, and Phillips does not govern this one, right?

[00:14:05] Speaker 04:
Your Honor, there was a very recent case, a presidential decision from this court, in Ray CSB International,

[00:14:14] Speaker 04:
that holds that patents as these have that have expired during the process are construed under the Phillips standard, even if they were addressed under the broadest reasonable interpretation standard before the board.

[00:14:27] Speaker 04:
The Enfish patents at issue here expired several months after the final written decisions were entered by the Patent Trial and Appeal Board.

[00:14:39] Speaker 04:
The court construed the phrase, a logical table,

[00:14:44] Speaker 04:
and emphasis on A. It said that the specification makes clear that the invention is directed to the arrangement of a single logical table.

[00:14:54] Speaker 04:
That construction affirmed the board's finding that Microsoft's invalidity proofs were inadequate if they did not contain all of the recited structures in a single table.

[00:15:07] Speaker 04:
So the court should affirm as to claims 32, 36, 42, and 43

[00:15:14] Speaker 04:
solely based upon the ENFISH 1 claim construction, because that agrees with the board's finding of the single table requirement.

[00:15:23] Speaker 04:
And that was the board's primary basis for finding that the invalidity proofs failed.

[00:15:29] Speaker 00:
Even if we agreed with you as to some of those claims, Microsoft is arguing that it has to claim 32, that they are pointing to a single table.

[00:15:40] Speaker 04:
Your Honor, I can only point to the board's fact findings on that issue.

[00:15:44] Speaker 04:
The board specifically, in its final written decisions, pointed to Dr. Hosking's, and this is Microsoft's expert, his admission, it's at A1361, in which the board concluded that Dr. Hosking admitted on cross-examination that he could not point to any single table in Chang that allowed determination of OIDs via text entry.

[00:16:08] Speaker 04:
The board also made another fact finding in its final written decisions, and this is at A1363,

[00:16:15] Speaker 04:
that Microsoft had a failure of proof on this issue because there were gaps in Chang's disclosure and Microsoft had never filled in the gaps adequately.

[00:16:25] Speaker 04:
Finally, briefly with respect to the searching claim construction issue, we would submit that the board simply responded to and addressed the arguments that Microsoft presented.

[00:16:40] Speaker 04:
All of the arguments presented concerning the determination of OIDs via text entry

[00:16:44] Speaker 04:
presented to the board and to this court are based on querying or searching.

[00:16:50] Speaker 04:
As this court observed in in Ray Baxter International, it said, unsurprisingly, the board only addressed Baxter's arguments and the others were waived.

[00:16:59] Speaker 04:
We have the same situation here, where the arguments that the board addressed were in the context of searching.

[00:17:06] Speaker 04:
There's no express holding by the board that searching is required.

[00:17:10] Speaker 03:
Can you address the indexing?

[00:17:13] Speaker 03:
claims what why why why is Microsoft wrong in saying based on what I think 1659 and so on Hosking's declaration that somebody skilled artisan reading visual basic would know I still have some work to do to implement this and what I would do is go and look around for something as fundamental as Sultan's manual to

[00:17:43] Speaker 03:
implement and that provides the motivation to go and look at Salton.

[00:17:51] Speaker 04:
The common sense argument that's articulated here was not squarely presented to the board.

[00:17:57] Speaker 04:
So the board did not consider a common sense argument.

[00:18:00] Speaker 04:
Second, the board made fact findings and the question for this court is whether there's substantial evidence to support those findings and it found specifically that Salton taught away

[00:18:11] Speaker 04:
In particular because of the fact-finding concerning concerning how it's expensive and rarely possible to fold decks I'm sorry to full text what you're referring to Your honor I don't have a a page site to the either the final written decision or the

[00:18:35] Speaker 04:
The briefing on that, I'll look for it and may be able to respond on that when I stand back up and rebuttal.

[00:18:41] Speaker 04:
I'd like to turn to the claim construction for object identification number.

[00:18:44] Speaker 04:
There's two important aspects of the object identification number in the appealed claims.

[00:18:52] Speaker 04:
First,

[00:18:53] Speaker 04:
The term is object identification number, or OID.

[00:18:57] Speaker 04:
And in every appealed claim, the claim states that the object identification number is to identify a row.

[00:19:05] Speaker 03:
Did you argue to the board that there was a difference between its use of the word define and your preferred term, identify?

[00:19:16] Speaker 04:
Your honor, we consistently argued that the construction should include identify and disagree with the board.

[00:19:23] Speaker 04:
There is not extensive briefing on the difference, but we certainly are on record explaining why identify is the correct term.

[00:19:35] Speaker 04:
The board, in our view, viewed define and identify as interchangeable terms akin to

[00:19:46] Speaker 04:
cease and desist or null and void.

[00:19:48] Speaker 04:
Viewing those terms as interchangeable would be error here for a number of reasons.

[00:19:54] Speaker 03:
Did you say to the board, oh, wait a minute, you're really quite wrong about that.

[00:19:58] Speaker 03:
These two terms are not interchangeable.

[00:20:03] Speaker 03:
We have among our, I forget what the term was you used, something like quibbles or tweaks or something like that after the institution decision that said we want to, we think you need to clarify, amend, elaborate.

[00:20:18] Speaker 03:
I don't recall reading anything in which you said one of the ways you should change this is to swap out the word define and put in the word identify.

[00:20:31] Speaker 04:
Your Honor, we urge that position by consistently pressing for our proposed construction all the way to the end of the process.

[00:20:41] Speaker 00:
Right, but your proposed construction, I mean, you had three different points, that uniqueness, immutable, and system generated.

[00:20:48] Speaker 00:
You even had separate sections in your brief to address those things.

[00:20:51] Speaker 00:
How does focusing on those things equate to swapping out, identify, and define?

[00:21:01] Speaker 04:
If you turn to claim 31 or claim 41, Your Honor, you'll see that the claims use object identification number to identify.

[00:21:09] Speaker 04:
They also refer to rows that define columns.

[00:21:15] Speaker 04:
The board, by construing OID as an array of bits that defines, effectively read out the identification limitation in the claims, not only impermissibly broadening the claims, but also as a practical matter breaking the structure

[00:21:31] Speaker 04:
of the claimed logical table.

[00:21:33] Speaker 04:
It can only be self-referential if the object identification number is an identifier that allows a row to define a column by sharing the same OID.

[00:21:45] Speaker 04:
And when the board misconstrued OID as a definer and removed the identifier limitation, the practical effect was that the board then looked anywhere in

[00:22:00] Speaker 04:
an alleged row defining a column in CIS columns, for example, said any information that would define that column could be an OID.

[00:22:11] Speaker 04:
Remember that these claims are computer operations.

[00:22:15] Speaker 04:
They're computer operation claims.

[00:22:18] Speaker 04:
And the question, as to Chang's anticipation, was whether the system had assigned a value as an identifier for a row or a column.

[00:22:27] Speaker 04:
Not whether any definitional information might be found somewhere in a row, but that was the effect of the board's construction.

[00:22:35] Speaker 04:
I'd also like to briefly touch on, before I leave, identify.

[00:22:39] Speaker 04:
I think it is important to note that Microsoft agrees that the construction of OID should include identify.

[00:22:48] Speaker 04:
They argue that it's harmless error that the board misconstrued the term.

[00:22:53] Speaker 04:
As I explained, not only because of the claims

[00:22:55] Speaker 04:
use, identify, and define, for example, in claims 31 and 41.

[00:23:01] Speaker 04:
Also because of the express claim language, taking out, identify, and replacing it with define dramatically changed the meaning of the claims.

[00:23:11] Speaker 04:
Microsoft also agrees with Enfish regarding the requirement that uniqueness should be an aspect of the construction of OID.

[00:23:23] Speaker 03:
How can it be the case, given, what is it, figure three, where there's a row and there's a column with the same OID, that every object has to have a unique OID?

[00:23:41] Speaker 03:
Your Honor, the figure- It's a column object and a row object.

[00:23:44] Speaker 03:
They have the same OID in the patent.

[00:23:49] Speaker 04:
Yes, and this is in the red brief at nine.

[00:23:51] Speaker 04:
You can find figure three there.

[00:23:54] Speaker 04:
Your Honor, our interpretation of figure three is that column 126 with OID 1019 and row 136, also with OID 1019, are the same database object.

[00:24:09] Speaker 04:
There is a unity or an identity there.

[00:24:12] Speaker 04:
They are identified by the same OID for the very important reason in these claims.

[00:24:18] Speaker 03:
As far as I recall, the patent does not actually say that.

[00:24:24] Speaker 03:
And the requirement in the claim is that the row with OID 1019 merely has to correspond to the column with OID 136 and 126.

[00:24:38] Speaker 03:
They don't actually have to be identical.

[00:24:43] Speaker 03:
And one would think that if they don't have to be identical, whatever else they are, they're not the same object.

[00:24:48] Speaker 03:
And the patent doesn't say, for these purposes, these are the same object.

[00:24:53] Speaker 04:
The words correspond are used.

[00:24:56] Speaker 04:
The effect from a technical standpoint is identity, because the same identifier is used in both places.

[00:25:04] Speaker 04:
You have identity.

[00:25:05] Speaker 04:
The board in the 559 case did address this dispute between Microsoft and Enfish.

[00:25:13] Speaker 04:
That's at A102 to 103.

[00:25:15] Speaker 04:
The board agreed with Enfish that to the extent such an OID is unique between the column and row.

[00:25:23] Speaker 04:
that we are persuaded that it's unique to every row and column, as Enfish asserts.

[00:25:28] Speaker 04:
It's also notable that Dr. Hosking agreed at A-98.

[00:25:34] Speaker 00:
Again, that agreement only relates to each individual column.

[00:25:39] Speaker 00:
It doesn't talk about database-wide, does it?

[00:25:45] Speaker 04:
Your honor, that's our interpretation.

[00:25:46] Speaker 04:
It does not speak to database-wide.

[00:25:48] Speaker 04:
However, I'd submit that because in Chang, neither the column name or column number would be uniquely represented in rows, that even under Microsoft's local uniqueness interpretation, none of Chang's disclosures would satisfy the uniqueness requirement as to the rows.

[00:26:08] Speaker 03:
Can I ask?

[00:26:09] Speaker 03:
I had, I guess, been under the impression of A103, that there's a typo there.

[00:26:15] Speaker 03:
that what the board was saying is we are not persuaded that it is unique to every row and column.

[00:26:20] Speaker 03:
How does this make sense otherwise?

[00:26:22] Speaker 03:
This is all about how they're rejecting your position of uniqueness as between rows and columns.

[00:26:34] Speaker 03:
Is there just not a missing knot in there?

[00:26:37] Speaker 04:
Your Honor, I didn't read it that way.

[00:26:39] Speaker 04:
We read it as the board agreeing with our interpretation

[00:26:45] Speaker 04:
The identity of the row column.

[00:26:47] Speaker 03:
Even though this is in the context of saying, look, the row and the column, they have the same OID, you cannot possibly have uniqueness in the sense you're arguing for.

[00:26:59] Speaker 04:
We didn't interpret it that way.

[00:27:01] Speaker 04:
I'll reserve my remaining time for a vote.

[00:27:04] Speaker 02:
Mr. Arman, you mentioned that you would provide a citation when you come back.

[00:27:10] Speaker 02:
When you come back, you can provide the citation, but no argument because it relates to the main appeal.

[00:27:21] Speaker 01:
If I could briefly, just to clarify and sharpen up on what our position is with respect to the construction of OID.

[00:27:30] Speaker 01:
We don't think that the board made a mistake that requires any adjustment by the panel here today.

[00:27:37] Speaker 01:
The construction of OID was the construction of just one term.

[00:27:42] Speaker 01:
It wasn't the construction of a whole phrase in any of the elements of the claim.

[00:27:47] Speaker 01:
So what the board did is they said OID means they looked at a portion of the patent that talked about OID's defining

[00:27:59] Speaker 01:
rows and columns and another portion of the patent that talked about OIDs being bid arrays and they put those two together and said for this term OID we're going to have this construction that the OIDs are bid arrays that define.

[00:28:17] Speaker 01:
The board didn't ignore the rest of the claim language.

[00:28:21] Speaker 01:
Each of the claims require that you have an OID that identifies rows and OIDs that identify columns.

[00:28:29] Speaker 01:
If we look at what the board said on A18 of its decision, they point out in the last sentence there that in fish's contentions are not commensurate with the scope of the claims because each of the independent claims recites that an OID is included, quote, to identify, close quote, each logical row and each logical column.

[00:28:53] Speaker 01:
So that the board was applying that identification requirement not ignoring it.

[00:28:59] Speaker 01:
If we go and look at how they applied it to Chang, the reference that anticipated claim 31 in the board's view at page A23, the first full sentence there at the top of the page says also Chang discloses in accordance with our construction that each database table includes column and row identifiers

[00:29:24] Speaker 01:
such as record ID, column numbers or column names, comporting with the proper interpretation of an OID as an array of bits that define a column.

[00:29:34] Speaker 01:
So we disagree with the idea that the board was ignoring the requirement that stated in the claim that an OID needs to identify.

[00:29:42] Speaker 01:
They applied it both in their claims construction analysis and also when they went to apply that construed language to,

[00:29:54] Speaker 01:
the prior art that was before them.

[00:29:56] Speaker 03:
Can I ask you this pesky question?

[00:29:58] Speaker 03:
Yes.

[00:29:58] Speaker 03:
Do you think that there's a missing knot in that sentence that says we are persuaded that it is unique?

[00:30:04] Speaker 01:
We do, Your Honor.

[00:30:06] Speaker 01:
The reason for that is based in part on what happened at the hearing.

[00:30:10] Speaker 01:
There was an extended discussion that took place between Council and the Board in which we were discussing column three, or figure three rather,

[00:30:21] Speaker 01:
and the fact that you have a column and a row that both have the same OID.

[00:30:25] Speaker 03:
If we look at... And they don't actually... Those two... The row and the column do not, according to the patent, actually have to have identical content, right?

[00:30:33] Speaker 01:
Correct.

[00:30:33] Speaker 01:
And they don't in the figure.

[00:30:35] Speaker 01:
Right.

[00:30:36] Speaker 01:
Okay.

[00:30:36] Speaker 01:
So we don't think that there is a way to have a universal uniqueness requirement and make sense of the intrinsic record.

[00:30:45] Speaker 01:
And with respect to the other things that Enfish is trying to add to the

[00:30:50] Speaker 01:
construction of OID that the board for example with the immutability requirement that they're proposing actually found that their extrinsic evidence wasn't persuasive.

[00:31:00] Speaker 01:
That's a finding that needs to be deferred to.

[00:31:02] Speaker 01:
With respect to system generation there's been a pivot on appeal to a disavowal argument and we would submit that there is not a basis for disavowal.

[00:31:15] Speaker 01:
I see that I'm

[00:31:16] Speaker 01:
I'm out of time.

[00:31:18] Speaker 02:
Thank you.

[00:31:20] Speaker 02:
Thank you, Mr. Campbell.

[00:31:21] Speaker 02:
Mr. Rahman can reply on the cross-appeal OID.

[00:31:27] Speaker 04:
Your Honor, first, the site I was unable to provide earlier, A156 to A157.

[00:31:33] Speaker 04:
Thank you.

[00:31:38] Speaker 04:
The board's error in confusing define and identify becomes apparent

[00:31:46] Speaker 04:
when you look at the final written decisions because the board's language reveals that although it uses the words identify and define, first of all, it said that it applied its construction of define in reaching its decisions.

[00:32:03] Speaker 04:
Second, and there's no finding by the board that Chang discloses assigning the column number or column name as the identifier for a row.

[00:32:15] Speaker 04:
Instead, that, of course, would have been required if a column number or column name was actually the system identifier for the row.

[00:32:25] Speaker 04:
And instead, the board found that that information was definitional information that could be found somewhere in a row in syscolumns.

[00:32:33] Speaker 04:
That definitional approach allowed the board to avoid the real question, which is whether the system identified that entire row

[00:32:42] Speaker 04:
with an OID that had been assigned based upon column name or number.

[00:32:47] Speaker 04:
That teaching's not present.

[00:32:52] Speaker 04:
And I see I'm out of time.

[00:32:54] Speaker 02:
Thank you, Mr. Rahman.

[00:32:56] Speaker 02:
We'll take the case under advisement.