`571-272-7822 Entered: April 15, 2014
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`13
`14
`15
`16
`17
`18
`19
`
`20
`21
`
`22
`
`23
`
`24
`
`25
`
`26
`
`27
`
`
`
`
`
`
`
`RECORD OF ORAL HEARING
`
` UNITED STATES PATENT AND TRADEMARK OFFICE
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
` ________________
`
`BEFORE THE PATENT TRIAL AND APPEAL BOARD
`
` ________________
`
` EMC Corporation
`
` Petitioner1
`
`
`
`
`
` V.
`
` Patent of Personal Web Technologies, LLC
`
`
`
` Patent Owner
`
`
`
`
`
`
`
` ________________
`
`
` Case IPR2013-00082 (Patent 5,978,791)
` Case IPR 2013-00083 (Patent 6,415,280)
` Case IPR 2013-00084 (Patent 7,945,544)
` Case IPR2013-00085 (Patent 7,945,539)
` Case IPR2013-00086 (Patent 7,949,662)
` Case IPR2013-00087 (Patent 8,001,096)
`
` ________________
`
` Record of Oral Hearing
`
`
`
`
`
`
`
`
`
`WITNESSES
`
`Before JONI CHANG, MICHAEL ZECHER and KEVIN TURNER (via video
`
`hookup), Administrative Patent Judges
`
`
`1 Note that the petitioners for IPR2013-00082 and IPR2013-
`00083 are EMC Corporation and VMware, Inc.
`
`
`
`
`Case IPR2013-00082 (Patent 5,978,791)
`Case IPR 2013-00083 (Patent 6,415,280)
`Case IPR 2013-00084 (Patent 7,945,544)
`Case IPR2013-00085 (Patent 7,945,539)
`Case IPR2013-00086 (Patent 7,949,662)
`Case IPR2013-00087 (Patent 8,001,096)
`
`
`APPEARANCES:
`
`ON BEHALF OF THE APPELLANT:
`
`JOSEPH A. RHOA
`
`Nixon & Vanderhye, P.C.
`
`901 North Glebe Road
`
`11th Floor
`
`Arlington, VA 22203
`
`
`
`
`
` P R O C E E D I N G S
`
`JUDGE CHANG: Good afternoon. Welcome to the
`
`Patent Trial and Appeal Board. This is the final hearing for
`
`six inter partes reviews, Case IPR2013-00082, IPR2013-00083,
`
`IPR2013-00084, IPR2013-00085, IPR2013-00086, and
`
`IPR2013-00087.
`
`The Board instituted these trials for six related
`
`patents owned by PersonalWeb Technologies on May 17, 2013.
`
`The transcript of this final oral-hearing will be entered in
`
`each of the cases, and it will be usable for all cases.
`
`At this time, we would like to have the counsel
`
`to introduce themselves, beginning with the Petitioner.
`
`MR. DICHIARA: Thank you, Your Honor. My name is
`
`Peter Dichiara, and we represent the Petitioners, EMC and
`
`
`
`2
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`
`
`Case IPR2013-00082 (Patent 5,978,791)
`Case IPR 2013-00083 (Patent 6,415,280)
`Case IPR 2013-00084 (Patent 7,945,544)
`Case IPR2013-00085 (Patent 7,945,539)
`Case IPR2013-00086 (Patent 7,949,662)
`Case IPR2013-00087 (Patent 8,001,096)
`
`
`VMware.
`
`With me I have my partners, Cindy Vreeland and
`
`David Cavanaugh, who will also represent the Petitioners.
`
`From the Petitioners we have Mr. Lynn from
`
`VMware, Mr. Brown, Mr. Gupta and Mr. Clark from EMC, and then
`
`we also have Dr. Clark from Princeton University.
`
`JUDGE CHANG: Welcome.
`
`MR. RHOA: Good afternoon, Your Honors, Joe Rhoa,
`
`on behalf of PersonalWeb, and with me is Mickey Gill, on
`
`behalf of PersonalWeb.
`
`JUDGE CHANG: Welcome. Thank you.
`
`Just want to remind the parties that from our
`
`last order, we said that each party has two hours, okay, to
`
`present their arguments.
`
`This is -- so Petitioner will proceed first,
`
`okay, to present its case as to the challenged claims and
`
`grounds the Board instituted for these six cases.
`
`Thereafter, the Patent Owner will respond to the Petitioner's
`
`case. And the Petitioner may reserve rebuttal time for its
`
`case. And we'll proceed this way, okay?
`
`And starting with the Petitioner.
`
`MR. DICHIARA: Thank you, Your Honor.
`
`
`
`3
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`
`
`Case IPR2013-00082 (Patent 5,978,791)
`Case IPR 2013-00083 (Patent 6,415,280)
`Case IPR 2013-00084 (Patent 7,945,544)
`Case IPR2013-00085 (Patent 7,945,539)
`Case IPR2013-00086 (Patent 7,949,662)
`Case IPR2013-00087 (Patent 8,001,096)
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`JUDGE CHANG: Do you have any handouts or
`
`demonstratives?
`
`MR. DICHIARA: I do.
`
`JUDGE CHANG: Can we have the copies of the
`
`slides?
`
`MR. DICHIARA: Okay.
`
`(Whereupon, there was a pause in the
`
`proceedings.)
`
`JUDGE CHANG: Please provide one for the court
`
`reporter.
`
`You may begin.
`
`MR. DICHIARA: Good afternoon. May it please the
`
`Board, as I mentioned before, I'm Peter Dichiara,
`
`representing the Petitioner in this case which is both EMC
`
`and VMware. EMC and VMware are the Petitioners in IPRs 82
`
`and 83, and for the remainder of the Petitioners, just EMC.
`
`And as the Board is aware, the PersonalWeb
`
`patents relate to unique data identifiers and that those
`
`identifiers are based on the contents of the data and which
`
`are used for basic file management functions.
`
`Petitioners have shown that these identifiers and
`
`the patents used with these identifiers are old. Indeed, the
`
`
`
`4
`
`
`
`Case IPR2013-00082 (Patent 5,978,791)
`Case IPR 2013-00083 (Patent 6,415,280)
`Case IPR 2013-00084 (Patent 7,945,544)
`Case IPR2013-00085 (Patent 7,945,539)
`Case IPR2013-00086 (Patent 7,949,662)
`Case IPR2013-00087 (Patent 8,001,096)
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`prior art not only meets all of the limitations of the
`
`challenged claims, but it operates just like the embodiments
`
`of the PersonalWeb patents. And here is the agenda we would
`
`like to follow this afternoon. We would like to do a brief
`
`overview of the PersonalWeb patents, followed by a very brief
`
`overview of the primary references, Woodhill, Langer and
`
`Kantor, and then go much more time to the disputed claim
`
`limitations. And as you had mentioned, we would like to
`
`reserve some time for rebuttal.
`
`JUDGE CHANG: Can you tell me how much time do you
`
`want in rebuttal?
`
`MR. DICHIARA: We're shooting for an hour.
`
`JUDGE CHANG: For an hour. Okay.
`
`JUDGE ZECHER: Quick question. The disputed claim
`
`limitations is the patent order there, the numbers, do they
`
`coincide with 82 through 7 chronologically there?
`
`MR. DICHIARA: Not exactly. Because we -- in the
`
`interest of efficiency, we, for example, we do the '87 case,
`
`the '096 after the '539 because the subject matter overlaps
`
`so much.
`
`But I will have very clear chapter slides saying
`
`what we are switching between, which patents.
`
`
`
`5
`
`
`
`Case IPR2013-00082 (Patent 5,978,791)
`Case IPR 2013-00083 (Patent 6,415,280)
`Case IPR 2013-00084 (Patent 7,945,544)
`Case IPR2013-00085 (Patent 7,945,539)
`Case IPR2013-00086 (Patent 7,949,662)
`Case IPR2013-00087 (Patent 8,001,096)
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`JUDGE ZECHER: Okay. That's all I ask, that you
`
`tell us which one you are referring to so we can look at the
`
`records.
`
`MR. DICHIARA: Yeah, I'll do my best.
`
`So, all of the challenged patents share a common
`
`patent specification and the PersonalWeb patents concerned
`
`data processing systems that process data, obviously.
`
`And, as the patents explain, the data item may be
`
`a file, may be a portion of a file, may be an object in an
`
`object-oriented programming -- program, or it can be in any
`
`other entity which can be represented by a sequence of bits.
`
`PersonalWeb patents, for example, describe processing data
`
`items that are either simple or made up of other data items
`
`and each data item has a corresponding unique identifier,
`
`which the patents typically refer to in the patent spec as
`
`the True Name.
`
`And as the patent summary explained, the identity
`
`or the identifier depends only on the data.
`
`The PersonalWeb patent describes something it
`
`refers to as a True File registry. And the True File
`
`registry is a database which stores the various True Names
`
`and the data items and other information about those data
`
`
`
`6
`
`
`
`Case IPR2013-00082 (Patent 5,978,791)
`Case IPR 2013-00083 (Patent 6,415,280)
`Case IPR 2013-00084 (Patent 7,945,544)
`Case IPR2013-00085 (Patent 7,945,539)
`Case IPR2013-00086 (Patent 7,949,662)
`Case IPR2013-00087 (Patent 8,001,096)
`
`
`items.
`
`And as we explained in our petitions, the patents
`
`use these identifiers to perform basic file management
`
`functions.
`
`Now, the inventors of the PersonalWeb patents
`
`thought that all the prior art used something known as
`
`context-based names.
`
`And a context-based name is more of a
`
`conventional name that you probably are familiar with.
`
`C:\mydirectory/myfile, things of that nature.
`
`And the inventors also thought that, in the prior
`
`art, there was no direct relationship between the data names
`
`and the data item.
`
`And, indeed, the inventors believe that they had
`
`invented content-based names. This was an argument that they
`
`emphasize in their patents themselves and it is an argument
`
`that they pursued during prosecution. And in this slide, for
`
`example, all of that emphasis is in the original. That's not
`
`stuff we put in there to embellish the slide.
`
`But the inventors were wrong, as we state in our
`
`petitions. Everyone can now agree that multiple prior art
`
`references had content-based names.
`
`
`
`7
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`
`
`Case IPR2013-00082 (Patent 5,978,791)
`Case IPR 2013-00083 (Patent 6,415,280)
`Case IPR 2013-00084 (Patent 7,945,544)
`Case IPR2013-00085 (Patent 7,945,539)
`Case IPR2013-00086 (Patent 7,949,662)
`Case IPR2013-00087 (Patent 8,001,096)
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`And the Board, in its decision to institute trial
`
`in these cases, has focused on three in particular as primary
`
`references: Woodhill, Langer and Kantor.
`
`And each of these references have the same
`
`content-based identifiers, have the same infrastructure and
`
`were used for the same basic file management functions.
`
`So, as I said earlier, I would like to briefly
`
`review the primary references, and I'll begin with Woodhill.
`
`And I will do this before diving into the disputed claim
`
`limitations.
`
`So, Woodhill, figure 1 is clear on each of the
`
`processes. And, in Woodhill, it processes data items such as
`
`files. Binary objects are something sometimes referred to as
`
`BOBs. And just like the PersonalWeb patents, Woodhill
`
`processes files of one or more binary objects.
`
`And just like the PersonalWeb patents, Woodhill
`
`identified binary objects with something it calls binary
`
`object identifiers or sometimes they'll use the acronym BOBID
`
`for short.
`
`Indeed, as it is right out of Woodhill, Woodhill
`
`recognizes that the critical feature is to create a binary
`
`object identifier based on the content of the data.
`
`
`
`8
`
`
`
`Case IPR2013-00082 (Patent 5,978,791)
`Case IPR 2013-00083 (Patent 6,415,280)
`Case IPR 2013-00084 (Patent 7,945,544)
`Case IPR2013-00085 (Patent 7,945,539)
`Case IPR2013-00086 (Patent 7,949,662)
`Case IPR2013-00087 (Patent 8,001,096)
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`Woodhill, like the True File registry, has
`
`something it refers to more simply as just the file database.
`
`The file database has the content-based identifiers, the
`
`binary object identifier shown in the slide along with other
`
`file information.
`
`And as we explained in our petitions, Woodhill
`
`uses the binary object identifiers for basic file management
`
`function such as identifying and accessing data, managing
`
`backups, comparing binary objects to recognize duplicates.
`
`And the next reference I would like to turn to is
`
`the Langer reference. Langer is a Usenet posting dated to
`
`12
`
`1981.
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`By way of background, Usenet was a network of
`
`computers used to disseminate information to something known
`
`as newsgroups. Langer was distributed to several newsgroups
`
`as you can see here in the header that's explored at
`
`alt.sources.d.comp.archives.admin.
`
`Langer was concerned with sharing content on the
`
`Internet as it then existed. This predates the Internet that
`
`we now know and love with the worldwide web and html browsers
`
`and the days when people used something called ftp or file
`
`transfer protocol.
`
`
`
`9
`
`
`
`Case IPR2013-00082 (Patent 5,978,791)
`Case IPR 2013-00083 (Patent 6,415,280)
`Case IPR 2013-00084 (Patent 7,945,544)
`Case IPR2013-00085 (Patent 7,945,539)
`Case IPR2013-00086 (Patent 7,949,662)
`Case IPR2013-00087 (Patent 8,001,096)
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`Langer processes data items such as files and
`
`packages of files. And just like the PersonalWeb patent, it
`
`identifies those files with MD5 hashes or MD5 codes. And,
`
`likewise, the packages, you know, identify those with MD5
`
`codes that are hash of hashes.
`
`Like the True File registry, Langer uses the
`
`unique identifiers with the central database. And they
`
`mentioned Archie and WAIS, for example. My slide is the
`
`instance where they were referring to Archie and used that
`
`database to associate the MD5 code with other information
`
`about the file such as its physical location.
`
`And just like the PersonalWeb patents used
`
`identifiers to access a file in the system, Langer uses an
`
`MD5 code to access the file in a network. And, indeed,
`
`PersonalWeb admits to this in its briefings.
`
`The next reference I would like to turn to is
`
`Kantor.
`
`Kantor discloses something it calls the FWKCS
`
`Contents Signature System. It's an acronym for the
`
`developer, Frederick W. Kantor, CS is the content signature.
`
`In particular, it concerns version 22 and it is used in BBS
`
`content. It's used in a BBS context to address the problem
`
`
`
`10
`
`
`
`Case IPR2013-00082 (Patent 5,978,791)
`Case IPR 2013-00083 (Patent 6,415,280)
`Case IPR 2013-00084 (Patent 7,945,544)
`Case IPR2013-00085 (Patent 7,945,539)
`Case IPR2013-00086 (Patent 7,949,662)
`Case IPR2013-00087 (Patent 8,001,096)
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`with duplicate files on a bulletin board service.
`
`Version 22 of Kantor dates back to August 10,
`
`1993.
`
`Kantor processes data items such as files and
`
`subject files just like the patents. Kantor identifies the
`
`files with a content-based identifier. It is a CRC hash and
`
`a link pinned together. In fact, they call it a content
`
`signature and it identifies zipfiles with a zipfile content
`
`signature, which is just like the patent with a hash of
`
`hashes. And just as the name suggests, depends on the
`
`contents of the data items.
`
`Like the True File registry, Kantor has something
`
`that it calls the CSLIST, which stands for content signature
`
`list. And the content signature list includes the content
`
`signatures and it includes other information about those
`
`files.
`
`And here is one of the excerpts from Kantor that
`
`shows the format of the content signature list.
`
`May I approach?
`
`And I'll just note, so the first row appears.
`
`What are those columns about? The first column is
`
`16_character_cs. It's the content signature. And underneath
`
`
`
`11
`
`
`
`Case IPR2013-00082 (Patent 5,978,791)
`Case IPR 2013-00083 (Patent 6,415,280)
`Case IPR 2013-00084 (Patent 7,945,544)
`Case IPR2013-00085 (Patent 7,945,539)
`Case IPR2013-00086 (Patent 7,949,662)
`Case IPR2013-00087 (Patent 8,001,096)
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`it, the row underneath, you'll see 32bitCRC32bitLen. That is
`
`a normal content signature. The CRC hash and the length
`
`appended.
`
`Underneath it you'll see CRCsum32LENSum32. It
`
`shows right here that this file content signature right at
`
`the hash of hashes of whatever inner files were in that
`
`zipfile along with other file information as you proceed left
`
`to right.
`
`Kantor uses the content signatures just like the
`
`patent does. The PersonalWeb patent uses for basic file
`
`management functions or partitions outlined in a number of
`
`these instances. Here is one such excerpt in which it talked
`
`about comparing a -- the file content signature with all the
`
`other files in an entire bulletin board service. And then
`
`Kantor explains it recognizes as redundant a zipfile is made
`
`up of pieces even if it is scattered throughout the system.
`
`And it can prevent wasteful duplication and protect against
`
`unwanted files.
`
`Kantor was published two separate ways. One was
`
`via a CD-ROM that was popular at the time. In particular, as
`
`our papers explained, there was one Walnut Creek, October of
`
`1993 CD-ROM.
`
`
`
`12
`
`
`
`Case IPR2013-00082 (Patent 5,978,791)
`Case IPR 2013-00083 (Patent 6,415,280)
`Case IPR 2013-00084 (Patent 7,945,544)
`Case IPR2013-00085 (Patent 7,945,539)
`Case IPR2013-00086 (Patent 7,949,662)
`Case IPR2013-00087 (Patent 8,001,096)
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`Here is a screen shot from our declarant, a
`
`gentleman by the name of Mr. Sadofsky, an Internet archivist.
`
`And he has entered a couple of declarations in this dispute,
`
`and here is one of them. You'll see a similar screen shot in
`
`paragraph five of Patent Owners Declarant Mr. Thompson. And
`
`as you'll see on the screen shot up in the menu bar, you'll
`
`see it say -- it might be a little tough to read,
`
`FWKCS122.ZIP. You'll also notice the timestamps, August
`
`10th, 1993.
`
`As the Board noted in its own decisions, Kantor
`
`speaks about another form of publication, namely from the
`
`bulletin board service called the Invention Factory. We had
`
`another declarant in this case, Mr. Sussil, who was the owner
`
`and system operator for the Invention Factory, and he too
`
`testified about this alternative means of publication and
`
`public accessibility, mainly the way to download the
`
`materials from the bulletin board service.
`
`So now that we've reviewed the patents quickly
`
`and the art quickly, we would like to turn to the disputed
`
`claim limitations and begin with '791 patent, which is
`
`IPR'82, and I would like to focus in particular on the claim
`
`limitations which are in dispute.
`
`
`
`13
`
`
`
`Case IPR2013-00082 (Patent 5,978,791)
`Case IPR 2013-00083 (Patent 6,415,280)
`Case IPR 2013-00084 (Patent 7,945,544)
`Case IPR2013-00085 (Patent 7,945,539)
`Case IPR2013-00086 (Patent 7,949,662)
`Case IPR2013-00087 (Patent 8,001,096)
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`As you're aware, Patent Owner has not disputed
`
`all the claims, at least independently. All the claim
`
`petitions, for example, they claim separately claims 29, 31
`
`or 32. And for several other claims there were claim
`
`limitations they didn't raise any dispute in their Patent
`
`Owner response. So, in the interest of time, I'd like to
`
`focus on the ones where there does seem to be a dispute.
`
`Woodhill is the primary reference.
`
`The Board correctly recognized that Woodhill
`
`satisfied the identity --
`
`JUDGE CHANG: Can I interrupt you?
`
`MR. DICHIARA: Yes.
`
`JUDGE CHANG: Okay. Before we get into the
`
`dispute claim limitations and Woodhill, can you talk about --
`
`a little bit about claim construction, because there is a
`
`claim constructions question --
`
`MR. DICHIARA: Certainly.
`
`JUDGE CHANG: -- regarding the Patent Owner's
`
`response.
`
`In their response, they offer a different claim
`
`construction than ours in the decision on institution, and we
`
`need to discuss function limitations identifying means for
`
`
`
`14
`
`
`
`Case IPR2013-00082 (Patent 5,978,791)
`Case IPR 2013-00083 (Patent 6,415,280)
`Case IPR 2013-00084 (Patent 7,945,544)
`Case IPR2013-00085 (Patent 7,945,539)
`Case IPR2013-00086 (Patent 7,949,662)
`Case IPR2013-00087 (Patent 8,001,096)
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`determining for any of a plurality of data items and so
`
`forth.
`
`Can you comment on that, whether you agree with
`
`the Patent Owner's proposed constructions now or not?
`
`MR. DICHIARA: We disagree.
`
`JUDGE CHANG: Okay.
`
`MR. DICHIARA: We think that the Board has
`
`correctly construed the term and construed it correctly under
`
`the procedures of inter partes review that brought us
`
`reasonable construction, and we think that's true for a
`
`couple different reasons.
`
`One, as the Board noted in its decision, the
`
`patent specifications certainly don't demand that there be
`
`compound data items. They could all be just regular old
`
`files.
`
`As it gets to the issue of whether to use
`
`something like an MD5 signature or something like that in
`
`particular, one of the items here on the screen addresses
`
`exactly that.
`
`I had the opportunity to cross-examine Patent
`
`Owners' expert, Dr. Dewar, and I asked him about the specific
`
`claim language which refers to identical data items will have
`
`
`
`15
`
`
`
`Case IPR2013-00082 (Patent 5,978,791)
`Case IPR 2013-00083 (Patent 6,415,280)
`Case IPR 2013-00084 (Patent 7,945,544)
`Case IPR2013-00085 (Patent 7,945,539)
`Case IPR2013-00086 (Patent 7,949,662)
`Case IPR2013-00087 (Patent 8,001,096)
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`the same identifier, and he admitted all hash functions, even
`
`trivial hash functions, satisfy that claim language. So, we
`
`maintained that the Board's construction is the broadest
`
`reasonable. We saw nothing in the Patent Owner's response
`
`which suggested that you went beyond reason or beyond
`
`broadest reasonable or why their proposal was the broadest
`
`reasonable. They want it, but they haven't explained why
`
`it's the broadest reasonable and why yours was too broad.
`
`JUDGE ZECHER: Can I follow-up on my questioning?
`
`MR. DICHIARA: Certainly.
`
`JUDGE ZECHER: Obviously, your construction that
`
`they proposed in their Patent Owner response has a few
`
`citations, one of which is column 12, line 54 through column
`
`13, line 19.
`
`Somewhat similar to what we cited when we
`
`construed the function limitations, we focused primarily on
`
`columns. Column 13, line 15 through 19 which talk about
`
`Hash(K) is an important basis for the competition. And you
`
`seem to indicate -- they seem to indicate that starting in
`
`column 12, that the function -- if we're looking at line 61,
`
`the function MP must have the following properties. Almost
`
`as if we need to read these properties into their
`
`
`
`16
`
`
`
`Case IPR2013-00082 (Patent 5,978,791)
`Case IPR 2013-00083 (Patent 6,415,280)
`Case IPR 2013-00084 (Patent 7,945,544)
`Case IPR2013-00085 (Patent 7,945,539)
`Case IPR2013-00086 (Patent 7,949,662)
`Case IPR2013-00087 (Patent 8,001,096)
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`construction of a treeing mechanism. How would you respond
`
`to that?
`
`MR. DICHIARA: I think that's not the broadest
`
`reasonable construction. I mean, there is support in this for
`
`that kind of calculation and there is support for broader as
`
`well as Dr. Dewar has admitted. So, I don't believe their
`
`construction is the broadest reasonable. I think it is
`
`purely driven, that after the decision in institute -- things
`
`weren't in their opening argument that it must be this, and
`
`it in must be that, and it must be the next thing, is a
`
`result of your decision where they're trying to narrow the
`
`construction at this point. Narrow the claimed scope to a
`
`claim constructions.
`
`JUDGE CHANG: I think we recognize that. This is a
`
`means plus function. So, under 35 USC 112, 6, you know, we
`
`have to look at the construction that restricts the
`
`specifications.
`
`MR. DICHIARA: Right.
`
`JUDGE CHANG: So, I guess I would like to know
`
`your five, on why isn't this in the hashline structure?
`
`MR. DICHIARA: 112,6 demands minimum corresponding
`
`structure, which you will see in all of your means plus
`
`
`
`17
`
`
`
`Case IPR2013-00082 (Patent 5,978,791)
`Case IPR 2013-00083 (Patent 6,415,280)
`Case IPR 2013-00084 (Patent 7,945,544)
`Case IPR2013-00085 (Patent 7,945,539)
`Case IPR2013-00086 (Patent 7,949,662)
`Case IPR2013-00087 (Patent 8,001,096)
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`function instructions where data processes a program a
`
`certain way. And, in this case, you would correctly identify
`
`that it is a data processor program to calculate the hash and
`
`then it provided those as examples.
`
`And I think that is the right construction. It
`
`could be that. It is not limited to that. SHA, for example,
`
`one of the ones you mentioned, the H is for hash. And so we
`
`believe that the broadest reasonable construction, which is
`
`supposed to govern this proceeding, is your construction is
`
`right and this is too narrow, their proposal is too narrow.
`
`JUDGE CHANG: Let me follow-up with one more
`
`question.
`
`There are five listed functions. So, are they the
`
`basic functions that any -- I heard you say that, but I just
`
`wanted to verify that that's what you said. Those five
`
`functions, is it always included in any MD algorithms?
`
`MR. DICHIARA: I can't answer it that broadly for
`
`any. I do know that the claim language when it says that
`
`identical data items will have the same identifier, that you
`
`don't need an MD5 for that. We confirmed that with their
`
`expert.
`
`
`
`The other characteristics listed in there, I'm
`
`18
`
`
`
`Case IPR2013-00082 (Patent 5,978,791)
`Case IPR 2013-00083 (Patent 6,415,280)
`Case IPR 2013-00084 (Patent 7,945,544)
`Case IPR2013-00085 (Patent 7,945,539)
`Case IPR2013-00086 (Patent 7,949,662)
`Case IPR2013-00087 (Patent 8,001,096)
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`sure that the MD5 satisfies that. I should say I believe
`
`that MD5 satisfies that. I don't know whether it's all MD5
`
`for whatever.
`
`JUDGE CHANG: It's the same.
`
`MR. DICHIARA: Same thing with SHA. The SHA, as I
`
`mentioned, the H is for secured hash algorithm, the same
`
`referring to hash N.
`
`JUDGE CHANG: Thank you.
`
`MR. DICHIARIA: So, I'll just proceed on to the
`
`next slide, because I think your questions address this.
`
`So, the next element that PersonalWeb disputes is
`
`the functional phrase and existence means and, in particular,
`
`the phrase that says for determining whether a particular
`
`data item is present in the system.
`
`And, PersonalWeb contends that Woodhill doesn't
`
`perform this function. And in fact, this is their longest,
`
`most repeated argument in their response. They devote the
`
`most pages to it. They repeat it the most times. And the
`
`Board correctly recognized that Woodhill satisfies this
`
`limitation relying on column 8. And I have that portion up
`
`here on the screen where Woodhill is explicit that it detects
`
`duplicate binary objects, that they can be recognized that
`
`
`
`19
`
`
`
`Case IPR2013-00082 (Patent 5,978,791)
`Case IPR 2013-00083 (Patent 6,415,280)
`Case IPR 2013-00084 (Patent 7,945,544)
`Case IPR2013-00085 (Patent 7,945,539)
`Case IPR2013-00086 (Patent 7,949,662)
`Case IPR2013-00087 (Patent 8,001,096)
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`the binary object identifiers are identical.
`
`And, on its face, this is showing that Woodhill
`
`determines whether a particular data item is present in the
`
`system, not once but twice. That's what duplicate means. It
`
`exists twice at two very particular locations. The Board was
`
`correct in relying on this passage for Woodhill.
`
`And then, we also have on this slide, there is
`
`more corroboration on this point.
`
`The next quote I have produced is from Woodhill
`
`column 9. This is the portion of Woodhill which is talking
`
`about a backup procedure where it maintains a BOB by BOB
`
`decision on whether to back one up to a backup device. It's
`
`trying to detect whether the BOB has changed or where it is
`
`something that's duplicate. And as the passage says, the
`
`binary object identifiers are calculated and then they're
`
`compared against their counterparts in the database. And
`
`this is just what the Board's decision of existence means.
`
`You'll go back and you'll see that the construction referred
`
`to step S 232 of the '791 patent, and that's what it says.
`
`It says to calculate the True Name before it in the database.
`
`So that's another piece of corroboration.
`
`The next item that I have up here is again some
`
`
`
`20
`
`
`
`Case IPR2013-00082 (Patent 5,978,791)
`Case IPR 2013-00083 (Patent 6,415,280)
`Case IPR 2013-00084 (Patent 7,945,544)
`Case IPR2013-00085 (Patent 7,945,539)
`Case IPR2013-00086 (Patent 7,949,662)
`Case IPR2013-00087 (Patent 8,001,096)
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`right out of our petition, and it's from the '791 patent.
`
`And it goes to the point that I raised earlier that Woodhill
`
`operates just like the embodiments of the patent. And in this
`
`passage you see the '791 patent says that the True Name of
`
`the file can be used to identify a file by contents. We
`
`talked about that earlier. No dispute on that. And then it
`
`continues.
`
`To confirm that the file matches its original
`
`contents or to compare two files. This is precisely what
`
`Woodhill is doing years earlier. It's comparing the current
`
`version of a file to a prior version of a file. Current
`
`version of a file to its original content.
`
`And I would invite you also to review Dr. Clark's
`
`direct testimony on this in this petition. At paragraph 87
`
`he noted other similarities between '791 patent and Woodhill.
`
`And, for example, he pointed out to column 36 of Woodhill,
`
`lines 42 to 45, where it is explicit -- strike that. I think
`
`I gave you -- said Woodhill's. It is column 36, lines 42 to
`
`45 of the '791 patent. And the '791 patent is explicit that
`
`it uses True Names for backups. That is exactly what Woodhill
`
`is doing. It's just like the embodiment in the patent.
`
`JUDGE ZECHER: Counsel, how did you respond to
`
`
`
`21
`
`
`
`Case IPR2013-00082 (Patent 5,978,791)
`Case IPR 2013-00083 (Patent 6,415,280)
`Case IPR 2013-00084 (Patent 7,945,544)
`Case IPR2013-00085 (Patent 7,945,539)
`Case IPR2013-00086 (Patent 7,949,662)
`Case IPR2013-00087 (Patent 8,001,096)
`
`
`1
`
`2
`
`3
`
`4
`
`5
`
`6
`
`7
`
`8
`
`9
`
`10
`
`11
`
`12
`
`13
`
`14
`
`15
`
`16
`
`17
`
`18
`
`19
`
`20
`
`21
`
`22
`
`Patent Owner's arguments that this comparison -- they seem to
`
`be arguing that it is only done on a single file whereas in
`
`their invention it does it on all of the files, or different
`
`files or any of the many files in the system.
`
`MR. DICHIARA: I would respond in a couple of
`
`ways.
`
`The first way I would respond is that Woodhill
`
`determines whether a particular date item is present in the
`
`system in a particularly efficient way. It is looking for
`
`duplicate content exactly where you would expect to find
`
`duplicate content. It's where you would expect to find
`
`duplicate content is in a prior version of a file. You might
`
`be editing a file, 99 percent of it is still what you had
`
`before. It is looking for it there.
`
`This is just what you would do if you were trying
`
`to construe a claim term and you looked for a claim term in
`
`the dictionary. If you were trying to construe the word
`
`widget, we would turn to a particular page, you would if
`
`widget's there. Maybe it's there, maybe it's not. You
`
`wouldn't go through every page in the dictionary to say its
`
`not there, just at the particular spot that you would expect
`
`to find it. So, I say Woodhill determines in a particularly
`
`
`
`22
`
`
`
`Case IPR2013-00082 (Patent 5,978,791)
`Case IPR 2013-00083 (Patent 6,415,280)
`Case IPR 2013-00084 (Patent 7,945,544)
`Case IPR2013-00085 (Patent 7,945,539)
`Case IPR2013-00086 (Patent 7,949,662)
`Case IPR2013-00087 (Patent 8,001,096)
`
`
`efficient way.
`
`The other thing I would like to say on that is
`
`that the Patent Owner, at a point, tries to criticize
`
`Woodhill saying it doesn't duplicate content in all the other
`
`instances and all the other spots in the system. And we've
`
`already heard, at least in the Eastern District of Texas
`
`case, t