`STATES PATENT AND TRADEMARK OFFICE
`
`Trial Nos.:
`
`IPR 2014~00059
`
`Re:
`
`US. Patent No. 6,415,280
`
`Patent Owners:
`
`PersonalWeb Technologies, LLC & Level 3 Communications
`
`Petitioner:
`
`Rackspace US, Inc. & Rackspace Hosting, Inc.
`
`DECLARATION OF ROBERT B. K. DEWAR
`
`1. My name is Robert B. K. Dewar.
`
`I reside at 1591 Carpenter Hill Road,
`
`Bennington, VT 05201, and am working for the software company Ada
`
`Core Technologies in New York, New York, that I founded.
`
`I have been
`
`retained by PersonalWeb Technologies, LLC.
`
`I understand that
`
`PersonalWeb is an owner of patents involved in several Inter Parte
`
`Reviews (IPRs). This declaration relates to U.S. Patent No. 6,415,280
`
`(the ‘280 patent), and the IPR that it is involved in which I understand to
`
`be IPR No. 2014—00059.
`
`2.
`
`I am currently Professor Emeritus at New York University in the
`
`Department of Computer Science. My curriculum vitae (CV) is attac
`
`
`
`
`
`at the rear of this declaration. My educational and professional
`
`background is summarized below.
`
`. To summarize my educational and academic background, I was educated
`
`at a preparatory school in England, and attended one year of public
`
`school there.
`
`I then immigrated to the United States where I attended the
`
`University of Chicago Lab School.
`
`I then attended the University of
`
`Chicago where I earned BS and PhD degrees in the field of Chemistry.
`
`My PhD work in Chemistry involved substantial use of computers. After
`
`receiving my PhD degree, I took a position as Assistant Professor of
`
`Computer Science at the Illinois Institute of Technology in 1968.
`
`.
`
`I thereafter received tenure as an Associate Professor of Computer
`
`Science at the Illinois Institute of Technology.
`
`I was Full Professor of
`
`Computer Science, CIMS, NYU, since 1975, and am currently Professor
`
`Emeritus at New York University in the Department of Computer
`
`Science.
`
`I was Chairman of the Department of Computer Science from
`
`1977-1980 at CIMS, NYU.
`
`I was Director of Undergraduate Studies,
`
`Computer Science from 1981—1987, CIMS, NYU.
`
`I was Associate
`
`Director of Courant Institute, from 1994—1997.
`
`. My teaching experience as a Professor includes (i) undergraduate courses
`
`relating to Introduction to programming for non-maj ors, first course for
`
`
`
`majors (CSl/CSZ equivalents, taught in Fortran, Pascal, SNOBOL, Ada
`
`95), Assembly language and Computer Architecture, Operating Systems,
`
`Microprocessors, Text processing systems, Programming Languages,
`
`Formal language theory, Numerical Analysis, Data processing, and
`
`Introduction to Algol-68, and (ii) graduate courses relating to Formal
`
`language theory, Theory of Computation, Algorithms, Programming
`
`Languages, Compilers, Data Communications, Microprocessor
`
`Architectures, Real Time Systems, Information Theory, and
`
`Programming by Transformation.
`
`.
`
`I have authored or co—authored numerous articles and books relating to
`
`computer science, programming, and data processing systems, as shown
`
`in my CV.
`
`. My research has centered on compilers, programming languages,
`
`operating systems, and microprocessor architectures.
`
`I have engaged in
`
`extensive consulting (apart from serving as an expert witness) in the area
`
`of operating systems and distributed systems, including writing several
`
`operating systems for Honeywell used in distributed applications such as
`
`airline reservations and freight management.
`
`.
`
`In 1994 I founded Ada Core Technologies to commercially exploit the
`
`work on Ada (programming language) that I was involved with at New
`
`
`
`York University.
`
`I became CEO and President of Ada Core
`
`Technologies.
`
`9.
`
`I retired from New York University in 2005.
`
`10.1 have received numerous awards including the SIGAda (Special Interest
`
`Group on Ada) award for outstanding Ada Community Contributions, as
`
`well as several other awards, which are listed in my attached CV.
`
`11. 1 am familiar with USENETs and Bulletin Board Systems, and used the
`
`same in the 1980s and 1990s.
`
`12.
`
`l have served as an expert witness on a number of occasions in cases
`
`involving patent and copyright issues. A list of several cases in which I
`
`served as an expert witness can be found in my attached CV.
`
`13. I am being compensated for my time at the rate of $500/hour for service
`
`that i am providing in connection with these inter Partes Reviews (IPRS).
`
`My compensation is not contingent upon my performance, the outcome
`
`of this review, or any issues involved or related to this review.
`
`14.
`
`l have reviewed and am familiar with at least the following documents,
`
`and any other document mentioned herein:
`
`the ‘280 patent; US. Patent
`
`No. 5,649,196 to Woodhill (Woodhill); Langer; the Mercer declaration;
`
`and the Mercer deposition transcript. It is my understanding that the “280
`
`patent has an effective filing date of April 11, 1995.
`
`
`
`My Understanding of the Law Regarding Patent Validity
`
`15. The following is what I have been told about the law regarding validity
`
`of a patent, and it represents my understanding of the same.
`
`It is my
`
`understanding that Rackspace (who 1 understanding is a “petitioner” in
`
`this proceeding) has the burden of proving invalidity in an IPR
`
`proceeding by a preponderance of the evidence. As explained below, it is
`
`my opinion that the challenged claims of the ‘280 patent are not invalid
`
`under this standard.
`
`16. Anticipation. It is my understanding that a claim of a patent is
`
`“anticipated” only if each and every element as set forth in the claim is
`
`found, either expressly or inherently described, in a single prior art
`
`reference. If even a single element is missing, the claim is not
`
`anticipated. It is my understanding that a feature is “inherent” in a
`
`reference only if that feature is necessarily present in the reference ~ not
`
`merely probably or possibly present. Furthermore, it is my understanding
`
`that in order to anticipate, a prior art reference must not only disclose all
`
`elements of the claim, but must also disclose those elements arranged as
`
`in the claim.
`
`17. Obviousness.
`
`It is my understanding that a patent claim is invalid for
`
`“obviousness” if the claimed invention as a whole would have been
`
`
`
`obvious to one of ordinary skill in the art at the time of the invention in
`
`view of a single prior art reference or a combination of prior art
`
`references.
`
`I understand that a determination of whether a claimed
`
`invention would have been obvious requires taking into consideration
`
`factors which include assessing the scope and content of the prior art, the
`
`differences between the prior art and the claimed invention, and the level
`
`of ordinary skill in the art. It is my understanding that the level of
`
`ordinary skill in the art is determined by considering the type of problems
`
`encountered in the art, the prior art solutions to those problems, the
`
`rapidity with which innovations are made, the sophistication of the
`
`technology involved, and the educational level of active workers in the
`
`field. It is my understanding that a prior art reference may be considered
`
`to teach away from the invention when a person of ordinary skill, upon
`
`reading the reference, would be led in a direction divergent from the path
`
`that was taken by the claimed invention; and that the general rule is that
`
`references that teach away cannot serve to create a prima facie case of
`
`obviousness.
`
`Level of Ordinary Skill in the Art
`
`18. As of 1995, it is my opinion that a person of ordinary skill in the art to
`
`which the ‘280 patent pertains would likely be a person with a bachelors
`
`
`
`degree in computer science with ten to fifteen years of teaching or work
`
`experience in the field. Of course, one of ordinary skill in the art might
`
`have either more formal education with less experience, or less formal
`
`education with more experience.
`
`Issues Involving the ‘280 Patent
`
`19. It is my understanding that the Board, in a decision dated April 15, 2014,
`
`instituted a trial regarding the ‘442 patent for only the following: (a)
`
`whether claims 10, 15, 16, 18, 25, 36 and 38 are anticipated under 35
`
`U.S.C. §102(e) by Woodhill (Ex. 1003); and (b) whether claims 10, 15,
`
`16, 18,25,31—33, 36 and 38 are obvious under 35 U.S.C. §103(a) over
`
`Woodhill (Ex. 1003) and Langer (Ex. 1004).
`
`Claim Constructions
`
`20.
`
`I have used the following interpretations of claim terms in. my
`
`declaration.
`
`I have been asked to use these constructions for purposes of
`
`this proceeding.
`
`21. “file” and “data file” (claims 10, 15, 18, 25, 36, 38). The ordinary
`
`meaning of “file” in this art at the time of the invention was the following
`
`definition given in the 1994 Microsoft Computer Dictionary: “file A
`
`complete, named collection of information, such as a program, a set of
`
`data used by a program, or a user-created document. A file is the basic
`
`
`
`unit of storage that enables a computer to distinguish one set of
`
`information from another. A file might or might not be stored in human—
`
`readable form, but it is still the ‘ Zue’ that binds a conglomeration of
`
`instructions, numbers, words or images into a coherent unit that a user
`
`can retrievez change, deletez save, or send to an output device.”
`
`(emphasis added) (Microsoft Press Computer Dictionary, Second Edition
`
`(1994) — EX. 2011.) This is consistent with a “named” data item referred
`
`to at col. 5:48-49 of the ‘280 patent. A “named” data item as described
`
`in the ‘280 patent means that the data item is a coherent unit that is
`
`individually manipulatable by a user so that a user can selectively
`
`retrieve, change, delete, save, or send it to an output device. In other
`
`words, one of skill in the art would have understood that a “named” data
`
`item is one that is necessarily a complete coherent unit that a user can
`
`selectively retrieve, change, delete, save, or send to an output device.
`
`Thus, the ordinary meaning of “file” as described in the 1994 Microsoft
`
`Computer Dictionary is both proper and consistent with the description of
`
`“file” in the ‘280 patent.
`
`22. A mere segment ofa basic file is NOT a “file.” Both the specification of
`
`the ‘280 patent and the 1994 Microsoft Computer Dictionary are clear in
`
`this respect. The specification of the ‘280 patent expressly difi’erentiaz‘es
`
`
`
`between a “file” and a mere “segment” of a file, making clear that they
`
`are different and that a mere “segment” is not a “file.” As shown in Fig.
`
`
`2 of the ‘280 patent, the patent differentiates between a “file” 120 and a
`
`
`mere “segment” 122 ofa basic file. (‘280 patent, Fig. 2; col 5:25—35.)
`
`The specification of the 6280 patent explains, as shown in Fig. 2, that a
`
`“file” is at a different position in the hierarchy than a mere “segment” of
`
`a file. (‘280 patent, col. 5:25-31.) The ‘280 patent also explains that
`
`while a “file” can be named by a user, a mere “segment” of a file cannot
`
`be so named. (‘280 patent, col. 5:44-45; col. 8:28.) Regardless ofhow
`
`“file” is construed, the specification (which includes the drawings) of the
`
`“280 patent makes clear that a mere “segment” cannot be a “file.” One of
`
`ordinary skill in the art, upon reviewing the specification of the ‘280
`
`patent, would have understood that files are different than segments — 126.,
`
`a mere segment cannot qualify as a “file.”
`
`23. Woodhill’s binary objects are “equivalent” to the “segments” in the ‘280
`
`patent, which demonstrates that Woodhill’s binary objects (and granules)
`
`are not “files.”
`
`24. According to the 1994 Microsoft Computer Dictionary, a mere segment
`
`of a file cannot be a “file” at least because: (a) a mere segment of a file is
`
`not a “complete” collection of information, (b) a mere segment of a file is
`
`
`
`not a coherent unit that a user can retrieve, change, delete, save or send to
`
`an output device, (0) a segment typically has no file header, and (d) a
`
`mere segment of a file is not named. (Microsoft Press Computer
`
`Dictionary, Second Edition (1994) - BX. 201 l.)
`
`25. I also note that this construction would NOT “preclude a data file from
`
`including smaller data files” in all cases. For example, a package such as
`
`a ZIP file may include a plurality of inner files — these would ALL be
`
`“files” under PO’s construction because each of these files would satisfy
`
`the limitations of a “file” discussed above. Each inner file would be
`
`individually named and each would satisfy the ordinary meaning of a
`
`“file” in the 1994 Microsoft Computer Dictionary discussed above. Both
`
`before formation of the ZIP file and after unpacking the ZIP file, each of
`
`the inner files thereof would have a file header and each would be
`
`individually manipulatable by users, with each being a coherent unit that
`
`a user can retrieve, change, delete, save or send to an output device. In
`
`contrast, that is not the case with the binary objects and granules in
`
`Woodhill. Binary objects and granules in Woodhill do not have headers,
`
`are not individually manipulatable as files by users, are not inner files of
`
`a package file (e.g., ZIP file), and do not satisfy the definition of a “file”
`
`in the 1994 Microsoft Computer Dictionary. A binary object or granule
`
`10
`
`
`
`in Woodhill: (a) is determined solely based on size (number of bits) and
`
`is NOT based on the extent of any coherent unit that is named or
`
`individually manipulatable; (b) has no header because it is based solely
`
`on size, (0) is NOT a coherent unit that a user can retrieve, change, delete,
`
`save or send to an output device, ((1) is NOT a “complete” collection of
`
`information, and (e) is not named. The binary objects and granules in
`
`Woodhill, for example, are mere portions of a basic file, do not have
`
`headers, cannot be named, and are not complete coherent units that a user
`
`can selectively change, delete, save or send to an output device.
`
`Petitioner’s expert admits, for instance, that a user cannot access a
`
`granule in Woodhill and then delete only that granule.
`
`26. One of ordinary skill in the art at the time of the invention would have
`
`also understood that a “file” has a file header. The 1994 Microsoft
`
`Computer Dictionary explains the ordinary meaning that “a file header
`
`identifies a data file by name, Size, and time and date of creation or
`
`revision; it might also identify the program with which the file was
`
`created.” (Microsoft Press Computer Dictionary, Second Edition (1994) —
`
`Ex. 2011.) Accordingly, a “file” at the time of the invention, and in the
`
`‘280 patent, should be construed as including a file header. This is
`
`important because binary objects and granules in Woodhill do not have
`
`ll
`
`
`
`file headers, and thus demonstrating that they cannot be “files” or “data
`
`files” for this additional reason.
`
`27. For example, consider a book that makes up a WORD “file.” The book
`
`includes many pages. The WORD file has a header. It would
`
`unreasonable to allege that an arbitrary portion of one page of the book
`
`constitutes its own “file.” In a similar manner, any allegation that a mere
`
`binary object or granule in Woodhill constitutes its own “file” is
`
`unreasonable, contrary to the ordinary meaning of “file” at the time of the
`
`invention, and contrary to the specification of the ‘280 patent which
`
`differentiates between files and segments.
`
`28. The proper construction for a “data file” is the same as the construction
`
`for a “file” discussed above, except that a “data file” includes data.
`
`29. Preambies ofciaims 36 and 36 are limiting. I have been instructed to
`
`assume that the preambles of claims 36 and 38 are limiting and entitled to
`
`patentable weight.
`
`30. “data identifier” (claims 10, 18, 25).
`
`It is my understanding that the
`
`PTAB construed “data identifier” as a “substantially unique data
`
`identifier for a particular element.” I have been instructed to assume this
`
`construction for this term for purposes of this proceeding.
`
`12
`
`
`
`3 l . “name”for a data file (claim 18). A “name” for a data file in this art is
`
`more than simply an identifier. It cannot reasonably be said that all
`
`identifiers are file names. For example, while a fingerprint is an
`
`identifier for a person, it is not a “name” for that person. Something is
`
`not a “name” for a file simply because it identifies a file. Any attempt to
`
`simply replace “name” with “identifier” in a construction would be
`
`incorrect and contrary to the claim language itself. Even Woodhill
`
`contrasts the difference between a file “name” and a binary object
`
`“identifier” (see e.g., file name 40 and file name 80 in Figs. 3—4 of
`
`Woodhill, compared to binary object identifier 74 in Fig. 3 of Woodhill),
`
`evidencing that they are not necessarily the same thing in this art. One of
`
`ordinary skill in the art at the time of the invention would have
`
`understood that there can be a significant difference between a file
`
`“name” and a mere “identifier.” One of ordinary skill in the art would
`
`have recognized that a “name” for a data file is used to both (a)
`
`particularly identify a particular file so that the name can be used to
`
`differentiate the file from other files, and (b) refer to that file, access that
`
`file, search for that file, and address that file so that a user can use the
`
`name to selectively retrieve, change, delete, save or send the file to an
`
`output device. (E.g., ‘280 patent, col. 16:13-16; col. 17:15—50; col. 19:1—
`
`13
`
`
`
`27; col, 23:61—67; col. 24:56—67; col. 34:35—49; and EX. 2011.) This is
`
`relevant because Woodhill’s binary object identifiers 74 and “contents
`
`identifiers?” are not “names” for data files.
`
`32. “a particular data file having a particular name specifying a location”
`
`(claim 18). This means that the particular data file has a name that
`
`particularly identifies the particular data file. And the “name” has the
`
`same construction as “name” above, except that the particular name also
`
`specifies a location (in addition to the other requirements of a “name”
`
`above). For example, see the “pathname” described in the specification
`
`of the ‘280 patent which contains both the name of a file and a location
`
`ofthat file. (E.g., ‘280 patent, col. 5:38—44; col. 8:27—28; col. 9:6—7; col.
`
`18:42—43.) This is important because Woodhill’s “file location 38” field
`
`is not a particular name for a data file because it does not particularly
`
`identify any data file.
`
`The Challenged Claims are Not Anticipated by Woodhill and Would
`
`Not Have Been Obvious
`
`33.
`
`It is my opinion that the challenged claims of the ‘280 patent are not
`
`anticipated by Woodhill and would not have been obvious for at least the
`
`following reasons.
`
`14
`
`
`
`Woodhill fails to disclose or suggest a data file having two
`
`different names as required by claim 18.
`
`34. Claim 18 of the ‘280 patent requires: “a particular data file having a
`
`particular name specifi/ing a location in the network at which the data
`
`file may be located .
`
`.
`
`. another name for the particular data file, the
`
`other name including a data identifier determined using a given function
`
`of the data, where said data used by the given function comprises the
`
`contents of the particular data file .
`
`.
`
`. ” Thus, claim 18 requires two
`
`different names for a particular data file.
`
`35. Both recited names must satisfy all requirements of a “name” for a data
`
`file as set forth above in the claim constructions section, including
`
`particularly identifying that data file. In additional to the basic
`
`requirements of a “name”, the first “name” recited in claim 18 must also
`
`specify a location for the particular data file, and the second “name”
`
`recited in claim 18 must be determined using a function of the data of the
`
`particular data file. Referring to the specification of the ‘280 patent, for
`
`example, the first “name” recited in claim 18 may be a “pathname” as
`
`described in the specification. (E.g., ‘280 patent, col. 5:38—44; col. 8:27—
`
`28; col. 9:6-7; col. 18:42-43.) A “pathname” is a name that contains both
`
`the name of a file and a location of that file. Id. For instance, the
`
`specification explains that a “pathname” may contain both a “processor
`
`15
`
`
`
`name” where the file is located and a “final file name” that was provided
`
`by user for particularly identifying the file. And the second “name”
`
`recited in claim 18 may be a True Name as described in the specification,
`
`which is determined using a function (e.g., MDS) of the data of the data
`
`file. Thus, the specification of the ‘280 patent fully supports two
`
`different names for a given file as recited in claim 18.
`
`36. It is my understanding that petitioner argues that the “data file” is a
`
`particular binary object and that the first recited name in claim 18
`
`corresponds to Woodhill’s “file location 38” as shown in Fig. 3 of
`
`Woodhill. This is incorrect. Woohill’s “file location 38” is not a “name”
`
`for a data file. Woodhill’s “file location 38” does not particularly
`
`identify any binary object or any granule. The allegation that “file
`
`location 38” field is a “name” for a data file is analogous to saying that
`
`“backup server 12” in Woodhill is a “name” for a file. This is illogical
`
`and unreasonable. There is simply no “name” for a binary object or
`
`granule in Woodhill that, in addition to particularly identifying a binary
`
`object or granule which a “name” must do, also specifies a location for
`
`the binary object or granule. Indeed, nothing in Record 34 of Woodhill
`
`identifies a particular binary object or granule. Claim 18 cannot be
`
`anticipated by Woodhill for at least these reasons, and there would have
`
`16
`
`
`
`been no logical reason to have modified Woodhill to have met the claim
`
`in this respect.
`
`37. i also note that neither binary objects nor granules in Woodhill are even
`
`“files.” Additionally, binary object identifiers 74 and “contents
`
`identifiers” in Woodhill are not file “names.”
`
`38. Woodhill discloses file name 40 in Fig. 3 and file name 80 in Fig. 4.
`
`(Woodhill, Figs. 3-4; col. 3:61; col. 4:56—57.) However, neither ofthese
`
`file names 40, 80 in Woodhill are “determined using a given function of
`
`the data, Where said data used by the given function comprises the
`
`contents of the particular data file,” and neither of these particularly
`
`identifies a binary object or granule. Other than 40 and 80, there is no
`
`other file name in Woodhill.
`
`Granules corresponding to “contents identifiers” in the alleged “update
`
`request” in Woodhill are never provided to the requesting local
`
`computer 20 in response to the “update request” (claims 10, 18, 25, 36
`
`and 38).
`
`39. Claims 10, l8, 25, 36 and 38 each require a “request” for a data file,
`
`which request includes the name (or identifier, or hash) of that particular
`
`data file. These claims further require that, in response to the request, the
`
`requested file is provided to the requesting computer or client. For
`
`example, claim 25 requires “in response to a client requestfor the
`
`particular data file, the request including at least the data identifier 0f
`
`17
`
`
`
`the particular datafile, providing the [requesting] client with the
`
`particular data file from a given one of the servers .
`
`. .” This subject
`
`matter cannot be met if it is alleged that “files” are granules and “names”
`
`are “contents identifiers” with respect to Woodhill’s granularization
`
`restore procedure.
`
`40. it is my understanding that petitioner relies upon Woodhill’s
`
`granularization restore procedure. However, in Woodhill’s
`
`granularization restore procedure granules corresponding to “contents
`
`identifiers” in the alleged request (“update request”) are NOT provided
`
`from the remote backup server 12 to the requesting computer in response
`
`to the “update request.” (Woodhill at col. 17:18—67.) Thus, Woodhill
`
`cannot meet these claims if the claimed data files are argued to be are
`
`granules and the names are argued to be “contents identifiers.” Moreover,
`
`no name of the large database file at issue is included in the “update
`
`request,” so that the claimed data file cannot be considered to be the large
`
`database file.
`
`41. In Woodhill, granules corresponding to “contents identifiers” in the
`
`alleged request (“update request”) are never provided to the requesting
`
`computer in response to the update request. (Woodhill at Fig. 51 and col.
`
`17:18-67.) The “contents identifiers” in the alleged “update request”
`
`18
`
`
`
`were calculated in step 444 and thus are not for granules that are being
`
`requested — the requesting local computer 20 already has the granules
`
`corresponding to the “contents identifiers” calculated in step 444 and
`
`therefore they are never provided to the local computer in response to the
`
`update request. Thus, the cited art fails to meet these claims.
`
`42. Turning to the “contents identifiers” for the reconstituted granules in
`
`Woodhill, the “contents identifiers” for the granules reconstituted at col.
`
`17:48—50 are only used for comparison in box 450 of Fig. 5i after the Fig.
`
`5h reconstitution of those granules, and are not included in any “request”
`
`regarding a data item.
`
`Binary objects corresponding to records 58 allegedly in the
`
`“update request” are not provided to the requesting
`
`computer in response to the update request (claims 10, 18, 25,
`
`36 and 38)
`
`43. In Woodhill, binary objects corresponding to the binary object
`
`identification records 58 argued to be in Woodhill’s “update request” are
`
`NOT provided to the requesting computer in response to the update
`
`request. (Woodhill at col. 17:18 to col. 18:9.) The granularization
`
`restore procedure in Woodhill transmits select granules — not binary
`
`objects. And record 58 and binary object identifier 74 do not identify
`
`particular granules. Thus, Woodhill cannot meet the above-identified
`
`l9
`
`
`
`claims if the claimed data files are argued to be binary objects and the
`
`identifiers are argued to be binary object identifiers 74 or binary object
`
`identification records 58.
`
`44. As explained above, claims 10, 18, 25, 36 and 38 each require a
`
`“request” for a data file, which request includes the name (or identifier,
`
`or hash) of that particular data file. These claims further require that, in
`
`response to the request, the requested file is provided to the requesting
`
`computer or client. For example, claim 25 requires “in response to a
`
`client requestfor the particular datafile, the request including at least
`
`the data identifier of the particular data file, providing the [requesting]
`
`client with the particular data file from a given one of the servers .
`
`. .”
`
`This subject matter cannot be met if the claimed “data file” is considered
`
`to be one of Woodhill’s binary objects and the'identifier to the binary
`
`object identifier 74 or record 58 related to that object, because the binary
`
`objects corresponding to the identifier allegedly in the “request” are not
`
`disclosed as being transmitted to the local computer 20 in response to the
`
`request in Woodhill’s granularization restore procedure.
`
`45. In Woodhill, binary objects corresponding to binary object identification
`
`records 58 argued to be in the request (“update request”) are NOT
`
`provided to the requesting computer in response to the request.
`
`20
`
`
`
`(Woodhill at Fig. 51 and col. 17:18—67.) Woodhill never discloses such
`
`an occurrence, and instead indicates to the contrary. The restore
`
`procedure transmits only particular granules based on whether they have
`
`changed — NOT binary objects.
`
`46. 1 also note that the granularization restore procedure in Woodhill is
`
`different than the self-audit procedure of Woodhill in that the different
`
`procedures transmit different items, use different comparisons, and so
`
`forth. The two procedures cannot be combined.
`
`47. Thus, in the granularization process upon which petitioner relies,
`
`Woodhill never provides any binary object to the local computer 20 in
`
`response to the “update request.” No binary object sequence of bits is
`
`ever sent from the backup file server 12 to a local computer 20 in the
`
`restore procedure of Woodhill’s granularization feature. In Woodhillis
`
`granularization process, the backup file server 12 sporadically sends only
`
`select “granules” of a large database file to the local computer.
`
`(Woodhill, col. 17:60-65.) Binary object identifiers 74 and records 58 do
`
`not identify granules, and certainly are not names or identifiers for
`
`granules. Woodhill’s granularization embodiment does not send large
`
`database files or binary objects from server 12 to the local computer —
`
`21
`
`
`
`just select granules. Thus, Woodhill fails to meet these claims for this
`
`reason as well.
`
`There is no request containing a binary ojbeet identifier 74 in
`
`Woodhill’s self-audit procedure (claims 10, 18, 25, 36 and 38)
`
`48. As explained above, claims 10, 18, 25, 36 and 38 each require a
`
`“request” for a data file, which request includes the name (or identifier,
`
`or hash) of that particular data file. These claims further require that, in
`
`response to the request, the requested file is provided to the requesting
`
`computer or client. For example, claim 25 requires “in response to a
`
`client requestfor the particular data file, the request including at least
`
`the data identifier ofthe particular data file, providing the [requesting]
`
`client with the particular data file from a given one ofthe servers .
`
`. .”
`
`49. Woodhill’s self—audit embodiment also fails to disclose or suggest this
`
`subject matter. (Woodhill, col. 18:10—38.) There is no “request” in
`
`Woodhill’s self—audit procedure that includes a binary object identifier 74
`
`or “contents identifier.” Woodhill explains in the description of the self-
`
`auditing procedure that the Record 58 is “stored in File Database 25.”
`
`(Woodhill, col. 18:16—19.) While Woodhill states that the Record 58 is
`
`“stored in File Database 25”, the Record 58 is never described as part of
`
`a “request” for the binary object. (Woodhill, col. 18:16-19.) Moreover,
`
`Woodhill elsewhere indicates that, in a message, portions of the Record
`
`22
`
`
`
`58 other than the binary object identifier 74 are used to identify a binary
`
`object. (Woodhill, col. 11:65 to col. 12:5.) Accordingly, in Woodhill’s
`
`self—auditing procedure there is no description of a “request” for a binary
`
`object or anything else that includes a binary object identifier 74. To the
`
`contrary, Woodhill indicates that any such “request” would NOT include
`
`a binary object identifier 74. (E.g., Woodhill, col. 11:65 to col. 12:5.)
`
`Woodhill indicates that binary object identifiers 74 are certainly not used
`
`for accessing binary objects. 1d. Additionally, there is no request for a
`
`“data file” in Woodhill’s self—audit procedure as explained herein.
`
`Woodhill restores a randomly selected binary object —— not a “data file.”
`
`(Woodhill, col. 18:25.) A binary object is not a “data file.” And neither
`
`a binary object identifier 74 nor a binary object identification record 58 is
`
`a “name” for a data file.
`
`Binary objects and granules are not “data files” as required
`
`by the challenged claims.
`
`50. The binary objects and granules in Woodhill are not “files” or “data
`
`files” for the reasons explained above in the claim construction section.
`
`Accordingly, the challenged claims cannot be met by the cited art for this
`
`reason as well. Petitioner relies on a “granule” or “binary object” from
`
`Woodhill as the alleged “data file” with respect to all challenged claims.
`
`Because these are not data files as explained above in the claim
`
`23
`
`
`
`construction section, petitioner’s allegations regarding all challenged
`
`claims necessarily fail.
`
`51. For instance, the challenged claims require a “request” for a data file that
`
`includes the name, identifier, or hash of that data file. Woodhill never
`
`requests a file. There are no requests in Woodhill for a large database file.
`
`52. Moreover, the challenged claims require names, identifiers, or hashes for
`
`data files to be used in particular manners. It is my understanding that
`
`petitioner relies solely on Woodhill’s “contents identifiers” for granules,
`
`and binary object identifiers 74 for binary objects, in this respect.
`
`However, neither “contents identifiers” nor binary object identifiers 74 in
`
`Woodhill identify any particular “file” in Woodhill. Again, binary
`
`objects and granules are not data files. Accordingly, if binary objects and
`
`granules are not “data files”, the petition fails in. many respects and all
`
`challenged claims must be deemed patentable.
`
`53.For example, claim 10 states as follows: “distributing a set of data
`
`files across a network of servers; determining a data identifierfor
`
`a particular data file, the data identifier being determined using a
`
`given function of the data, wherein said data used by the given
`
`function to determine the data identifier comprises the contents of
`
`the particular data file; and in response to a requestfor the
`
`24
`
`
`
`particular data file, the request including at least the data
`
`identifier of the