`Case 3:17-cv-05659-WHA Document 371-15 Filed 02/14/19 Page 1 of 7
`
`EXHIBIT 11
`EXHIBIT 11
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`
`Case 3:17-cv-05659-WHA Document 371-15 Filed 02/14/19 Page 2 of 7
`
`Page 5
`
`·1· · · ·Networks.
`·2· · · · · · · · MS. HEDVAT:· Shannon Hedvat from
`·3· · · ·Kramer Levin Naftalis & Frankel on behalf of
`·4· · · ·the Plaintiff, Finjan, Inc., and the
`·5· · · ·witness.
`·6· · · · · · · · · · · · · *· * *
`·7· · · · · · · · · MICHAEL D. MITZENMACHER, a
`·8· · · ·witness called for examination by counsel
`·9· · · ·for the Defendant, having been
`10· · · ·satisfactorily identified by the reporter,
`11· · · ·being first sworn by the Notary Public, was
`12· · · ·examined and testified as follows:
`13· · · · · · · · · · · · · *· * *
`14· · · · · · · · · · · · EXAMINATION
`15· · · ·BY MS. CARSON:
`16· ·Q.· Good morning.· Could you please state your
`17· · · ·name and spell it for the record.
`18· ·A.· Michael Mitzenmacher, M-I-C-H-A-E-L.
`19· · · ·Mitzenmacher is M-I-T-Z-E-N-M-A-C-H-E-R.
`20· ·Q.· Do you prefer to be referred to by doctor?
`21· ·A.· Whatever is convenience for you, I suppose.
`22· ·Q.· Okay.· You've been deposed before, correct?
`23· ·A.· Yes.
`24· ·Q.· Do you know how many times?
`
`Page 6
`
`·1· ·A.· No.
`·2· ·Q.· Approximately?
`·3· ·A.· Probably more than 25 at this point.
`·4· ·Q.· You understand you've just taken an oath to
`·5· · · ·tell the truth?
`·6· ·A.· Yes.
`·7· ·Q.· And you understand that that oath has the
`·8· · · ·same force and effect as if given in a court
`·9· · · ·of law before a judge or a jury, correct?
`10· ·A.· I would assume so, sure.
`11· ·Q.· Is there any reason you cannot give me your
`12· · · ·best testimony today?
`13· ·A.· Not that I am aware of.
`14· ·Q.· What did you do to prepare for your
`15· · · ·deposition today?
`16· ·A.· I read over some of the various documents,
`17· · · ·in particular my report.· I met briefly with
`18· · · ·counsel yesterday.
`19· ·Q.· Which counsel?
`20· ·A.· Counsel here, Shannon.
`21· ·Q.· Did you do anything else?
`22· · · · · · · · MS. HEDVAT:· I caution the witness
`23· · · ·to not divulge anything that's subject to
`24· · · ·privilege.
`
`Page 7
`
`·1· ·A.· Not that I can recall.· Again, it was
`·2· · · ·meeting and I went over various documents.
`·3· ·Q.· When were you retained for this matter?
`·4· ·A.· I would have to go back and look.· I mean,
`·5· · · ·they suggested this was upcoming some time
`·6· · · ·ago but I don't think I was officially
`·7· · · ·retained until I think it was just a few
`·8· · · ·months ago or when I first put in the hours.
`·9· ·Q.· And you're being paid for the time spent on
`10· · · ·this case, correct?
`11· · · · · · · · MS. HEDVAT:· Objection, form.
`12· ·A.· Yes.
`13· ·Q.· How much are you being paid?
`14· ·A.· $750 an hour.
`15· ·Q.· Is that your standard rate?
`16· · · · · · · · MS. HEDVAT:· Objection, form.
`17· ·A.· Yes.· That's within my standard rates.
`18· ·Q.· What do you mean by that's within your
`19· · · ·standard rates?
`20· ·A.· Some clients vary by small amounts, but 750
`21· · · ·is the rate I've been charging Finjan for a
`22· · · ·while and I've kept it at that rate.
`23· ·Q.· What is the range of standard rates you
`24· · · ·charge clients?
`
`Page 8
`
`·1· · · · · · · · MS. HEDVAT:· Objection, form.
`·2· ·A.· I believe it's currently at $750 an hour to
`·3· · · ·$875 an hour.
`·4· ·Q.· Do you know how much time you've spent on
`·5· · · ·this case?
`·6· ·A.· Not exactly.
`·7· ·Q.· Do you know approximately how much time
`·8· · · ·you've spent on this case?
`·9· ·A.· I would have to go back and look but it's
`10· · · ·probably nearing 100 hours.
`11· ·Q.· So you've spent somewhere in the range of
`12· · · ·nearing ten hours in the past couple months
`13· · · ·on this case?
`14· · · · · · · · MS. HEDVAT:· Objection, form,
`15· · · ·misstates testimony.
`16· ·A.· Yeah.· I'd have to go back and check but my
`17· · · ·guess is it's more than 50, less than 150,
`18· · · ·so somewhere within that range.
`19· ·Q.· You've been an expert for Finjan before,
`20· · · ·correct?
`21· ·A.· Yes.
`22· ·Q.· In how many matters?
`23· ·A.· I guess I'd have to go back and count.· I've
`24· · · ·created a sheet of cases in the last five
`
`
`
`Case 3:17-cv-05659-WHA Document 371-15 Filed 02/14/19 Page 3 of 7
`
`Page 37
`
`·1· ·A.· Yes.
`·2· ·Q.· And that hashing function existed well
`·3· · · ·before the '780 Patent as well, correct?
`·4· · · · · · · · MS. HEDVAT:· Objection, form.
`·5· ·A.· I would have to say that I would have to go
`·6· · · ·back and look up the dates but I believe
`·7· · · ·that might be true.
`·8· ·Q.· Within the context of Claim 1 of the '780
`·9· · · ·Patent, is it limited to any particular
`10· · · ·hashing function?
`11· · · · · · · · MS. HEDVAT:· Objection, form.
`12· ·A.· Again, when we discuss limitations, I'm
`13· · · ·aware that sounds more like validity
`14· · · ·invalidity issues than infringement issues.
`15· · · ·I've been focused on infringement issues,
`16· · · ·like the hash functions that were being used
`17· · · ·in the context as you've pointed out.· Hash
`18· · · ·functions like MD5 and SHA-256 that arise in
`19· · · ·my report are known hash functions. I
`20· · · ·don't -- again, I haven't considered what
`21· · · ·would require some sort of specific
`22· · · ·limitations on the possible range of hashing
`23· · · ·functions.· That would be something I'd have
`24· · · ·to think more about.
`
`Page 38
`·1· ·Q.· You opined on claim construction issues in
`·2· · · ·this case, correct?
`·3· ·A.· Yes.
`·4· ·Q.· And opining on the meaning of the claims,
`·5· · · ·did you understand that hashing function was
`·6· · · ·limited to any particular type of hashing
`·7· · · ·function?
`·8· · · · · · · · MS. HEDVAT:· Objection, form.
`·9· ·A.· So I believe in paragraphs 39 to 41 is where
`10· · · ·the issue of claim constructions related to
`11· · · ·hash functions arise.· I certainly discuss
`12· · · ·things that are not limitations.· Again in
`13· · · ·the context I think of paragraph 39 in
`14· · · ·particular, whether there might be other
`15· · · ·limitations, you know, again, I don't see
`16· · · ·anything specific in the claim language and
`17· · · ·I can't recall anything in the specification
`18· · · ·if there are such limitations, I don't think
`19· · · ·they're affected by infringement analysis.
`20· ·Q.· So in so far as you were doing your
`21· · · ·infringement analysis, you just checked to
`22· · · ·see if the Juniper products performed a
`23· · · ·hashing function, and there wasn't anything
`24· · · ·more specific about that so it could have
`
`Page 39
`·1· · · ·been an MD5, a SHA-256?· That didn't matter?
`·2· · · ·Any type of hashing function in your
`·3· · · ·infringement analysis would have satisfied
`·4· · · ·the claim; is that fair?
`·5· · · · · · · · MS. HEDVAT:· Objection, form,
`·6· · · ·misstates testimony.
`·7· ·A.· I don't think that's what I said.· What I
`·8· · · ·said is when I look at infringement, I look
`·9· · · ·for things.· The claim element calls for a
`10· · · ·hashing function so I would look for a
`11· · · ·hashing function.· Here the sort of hashing
`12· · · ·functions that are being used are sort of
`13· · · ·well-known hashing functions, including MD5,
`14· · · ·SHA-256 or combinations thereof, and so all
`15· · · ·of those clearly fit within the standards as
`16· · · ·you pointed out.· They're sort of known hash
`17· · · ·functions.
`18· · · · · · ·So to the extent that I dealt with
`19· · · ·issues regarding limitations, I believe it's
`20· · · ·discussed in paragraph 39 that there was, I
`21· · · ·think, a response to some of the comments by
`22· · · ·Rubin but I don't think there are
`23· · · ·limitations of the order that he described.
`24· ·Q.· When you were performing your infringement
`
`Page 40
`·1· · · ·analysis, did you have an understanding as
`·2· · · ·to whether Claim 1 requires that you fetch
`·3· · · ·the software components identified by the
`·4· · · ·one or more references before you perform a
`·5· · · ·hashing function?
`·6· · · · · · · · MS. HEDVAT:· Objection, form.
`·7· ·A.· I think the claim language says that you
`·8· · · ·perform a hashing function on the
`·9· · · ·Downloadable and the fetch software
`10· · · ·components to generate a Downloadable ID.
`11· · · ·So that was the phrasing and understanding
`12· · · ·used in my analysis.· There are certainly
`13· · · ·ways that you could perform a hashing
`14· · · ·function on the Downloadable and the fetch
`15· · · ·software components that would have
`16· · · ·different tempo considerations depending on
`17· · · ·the structure of the hashing function and
`18· · · ·the timing of the system.
`19· ·Q.· So when you were performing your
`20· · · ·infringement analysis, was it your
`21· · · ·understanding that a system that started
`22· · · ·performing a hashing function before it
`23· · · ·completed the fetching element could
`24· · · ·infringe?
`
`
`
`Case 3:17-cv-05659-WHA Document 371-15 Filed 02/14/19 Page 4 of 7
`
`Page 41
`
`·1· · · · · · · · MS. HEDVAT:· Objection, form.
`·2· ·A.· Certainly you were required to perform the
`·3· · · ·hashing function on the Downloadable and the
`·4· · · ·fetch software components to generate the
`·5· · · ·Downloadable ID.· The actual timing of when
`·6· · · ·various parts of that operation may occur I
`·7· · · ·don't believe is specified or a requirement
`·8· · · ·in the claim.
`·9· ·Q.· Is there anything in Claim 1 that recites a
`10· · · ·requirement that you have to store the
`11· · · ·Downloadable ID?
`12· ·A.· Do you mind if I look at the patent?
`13· ·Q.· Sure.
`14· · · · · · · · (US Patent No. 6,804,780 marked
`15· · · ·Exhibit No. 1038 for Identification.)
`16· ·Q.· The patent's been marked as Exhibit 1038.
`17· ·A.· I'd say that may be a legal question. I
`18· · · ·don't see any specific language in the claim
`19· · · ·as I look at now referring to restoring,
`20· · · ·but, again, when I was looking from the
`21· · · ·aspect of infringement, as I believe I
`22· · · ·mentioned in the declaration, in this case
`23· · · ·the Downloadable ID is stored and typically
`24· · · ·that would be a use you would generate the
`
`Page 42
`·1· · · ·Downloadable ID and store it for later use,
`·2· · · ·but I don't see any specific language.
`·3· · · ·Whether that was implicit might be a legal
`·4· · · ·question but I don't think it would affect
`·5· · · ·my legal analysis since it's stored in this
`·6· · · ·case.
`·7· ·Q.· Prior to the '780 Patent it was known you
`·8· · · ·could hash an executable file, correct?
`·9· · · · · · · · MS. HEDVAT:· Objection, form.
`10· ·A.· I'd say generally it was known that you
`11· · · ·could hash data in various forms which would
`12· · · ·include potentially a single executable
`13· · · ·file.
`14· ·Q.· Is it fair to say that prior to the '780
`15· · · ·Patent one method of virus detection was to
`16· · · ·hash the file and compare the hash to a list
`17· · · ·of known hashes that were malware?
`18· · · · · · · · MS. HEDVAT:· Objection, form.
`19· ·A.· So I'd say that may have been an approach
`20· · · ·used for malware detection.· I would say the
`21· · · ·exact timing of when the hashing of single
`22· · · ·files for any sort of malware detection when
`23· · · ·that started, I'd have to go back and look
`24· · · ·specifically.· Again, that's different than
`
`Page 43
`
`·1· · · ·the requirements of Claim 1 so I didn't
`·2· · · ·focus on when those specific dates were for
`·3· · · ·for that type of processing in my analysis.
`·4· ·Q.· Does anything in Claim 1 limit it to the
`·5· · · ·context of dynamic analysis?
`·6· · · · · · · · MS. HEDVAT:· Objection, form.
`·7· ·A.· So can you clarify what you mean by dynamic
`·8· · · ·analysis in this context?
`·9· ·Q.· Does dynamic analysis have a meaning to you
`10· · · ·as a network security expert?
`11· ·A.· I would say an issue is that dynamic
`12· · · ·analysis is a general enough term that it's
`13· · · ·used in multiple contexts, so that's why I
`14· · · ·asked for the clarification in that it can
`15· · · ·mean different things depending on the
`16· · · ·setting even within the limited confines of
`17· · · ·network security.· So that's why I asked if
`18· · · ·you could clarify what you mean by dynamic
`19· · · ·analysis in this context.
`20· ·Q.· So when someone says to you "dynamic
`21· · · ·analysis" in the context of computer
`22· · · ·security, can you provide me with what that
`23· · · ·means to you?
`24· · · · · · · · MS. HEDVAT:· Objection, form, asked
`
`Page 44
`
`·1· · · ·and answered.
`·2· ·A.· Again, it can mean different things
`·3· · · ·depending on the context.· I could provide
`·4· · · ·one possible method or way in which it's
`·5· · · ·used but that would not be limiting.
`·6· ·Q.· Sure.· Go ahead.
`·7· ·A.· Okay.· So one way dynamic analysis would be
`·8· · · ·used would be to say examine a running
`·9· · · ·program that a static analysis in that case
`10· · · ·would be to take a code object, say in its
`11· · · ·binary form, and simply examine the actual
`12· · · ·binary without examining it without running
`13· · · ·their functioning, whereas a dynamic
`14· · · ·analysis would be not just simply examining
`15· · · ·the code as a static object but watching it
`16· · · ·perform as it runs.· I'd like to specify
`17· · · ·that that's just one way in which I think
`18· · · ·the term is used.
`19· ·Q.· Understood.· So using what you've just
`20· · · ·described as the meaning for dynamic
`21· · · ·analysis, is there anything in Claim 1 that
`22· · · ·limits the claim to dynamic analysis
`23· · · ·context?
`24· · · · · · · · MS. HEDVAT:· Objection, form,
`
`
`
`Case 3:17-cv-05659-WHA Document 371-15 Filed 02/14/19 Page 5 of 7
`
`Page 65
`·1· · · ·sort of at a minimum what I can think of off
`·2· · · ·the top of my head.
`·3· ·Q.· Does a Microsoft Word document that does not
`·4· · · ·contain a macro have executable code?
`·5· ·A.· In the setting that I've described, I would
`·6· · · ·interpret it that way, although I don't
`·7· · · ·think that affects or impacts my
`·8· · · ·infringement analysis.
`·9· ·Q.· Is it your understanding that something is a
`10· · · ·Downloadable so long as the file type could
`11· · · ·contain executable code?
`12· ·A.· I think I would say that issue doesn't arise
`13· · · ·or doesn't come up in my infringement
`14· · · ·analysis so I may have to consider it more
`15· · · ·carefully.· And I believe that relates to
`16· · · ·the answers I sort of previously provided.
`17· · · ·Again, I believe there is a default in the
`18· · · ·setting of security.· The appropriate action
`19· · · ·is to treat everything as a danger.· And, in
`20· · · ·particular, one of the issues is sometimes
`21· · · ·we don't know in advance or recognize what
`22· · · ·the dangers are.· So, again, from my
`23· · · ·standpoint, I would view such document
`24· · · ·files, even if they didn't contain macros,
`
`Page 66
`·1· · · ·as being Downloadables in that they provide
`·2· · · ·instructions to the computers and some of
`·3· · · ·those instructions may be -- may yield a
`·4· · · ·threat that we don't understand in advance.
`·5· ·Q.· Is there any type of file that does not
`·6· · · ·provide instructions to the computer?
`·7· · · · · · · · MS. HEDVAT:· Objection, form.
`·8· ·A.· Yes.· I would say in various contexts you
`·9· · · ·would think of documents that do not provide
`10· · · ·instructions to the computer.
`11· ·Q.· Like what?
`12· ·A.· So, again, I think we've talked about
`13· · · ·contexts where you would not think of text
`14· · · ·files as providing instructions.· You may
`15· · · ·have to take care in other situations where
`16· · · ·they might lead to executable instructions
`17· · · ·that you'd have to be aware of them, but
`18· · · ·they're definitely context or situations
`19· · · ·where I don't think you'd view text files as
`20· · · ·threats or as possible executables.
`21· ·Q.· So would a text file be a Downloadable
`22· · · ·within the meaning of Claim 1?
`23· ·A.· As I answered before, I would say typically
`24· · · ·no, although again you may have to have some
`
`Page 67
`·1· · · ·care in that a text file could conceivably
`·2· · · ·contain computer code so there may be
`·3· · · ·context where if the text file can be used
`·4· · · ·or treated as an executable application that
`·5· · · ·you may have to be aware of it, but in a
`·6· · · ·typical instance or setting, you would not
`·7· · · ·necessarily think of a text file as a
`·8· · · ·Downloadable.
`·9· ·Q.· Can you think of any other types of files
`10· · · ·that do not provide instructions to a
`11· · · ·computer?
`12· ·A.· Something typically that I think of?· Files.
`13· · · ·They're either text files or they're
`14· · · ·associated with a program or a binary
`15· · · ·themselves.· So I'm trying to expand or go
`16· · · ·through the list of sorts of files that
`17· · · ·would not be considered one of those two
`18· · · ·things.
`19· · · · · · ·I think there are variations on text
`20· · · ·files.· For instance, you can look at
`21· · · ·compressed forms of text files, such as JSON
`22· · · ·files which would fall into sort of the same
`23· · · ·class.· I'd say also, depending on the
`24· · · ·context, there would be various sorts of
`
`Page 68
`·1· · · ·data files where, again, typically you might
`·2· · · ·not consider them executables so they would
`·3· · · ·not be Downloadables, although I'd provide
`·4· · · ·the same caveat I provide with text files
`·5· · · ·which is that the separation between data
`·6· · · ·and instructions in computer systems is a
`·7· · · ·tentative one that can depend on context.
`·8· ·Q.· Can you think of any type of file that would
`·9· · · ·never be considered a Downloadable?
`10· · · · · · · · MS. HEDVAT:· Objection, form.
`11· ·A.· I would say that's context dependent.
`12· · · ·Again, in most instances there are various
`13· · · ·forms of text files or other data files that
`14· · · ·might have, for instance, a different
`15· · · ·extension but would correspond to data files
`16· · · ·that might contain text or binary
`17· · · ·information that you would typically not
`18· · · ·consider to be Downloadables.· As always,
`19· · · ·you need to consider or examine the context
`20· · · ·to see how they might be being used.
`21· ·Q.· So I'm just trying to figure out whether
`22· · · ·there's anything -- regardless of context,
`23· · · ·right, is there anything that you're willing
`24· · · ·to say would never be a Downloadable?· Any
`
`
`
`Case 3:17-cv-05659-WHA Document 371-15 Filed 02/14/19 Page 6 of 7
`
`Page 69
`
`·1· · · ·type of file?
`·2· · · · · · · · MS. HEDVAT:· Objection, form.
`·3· ·A.· I would say I haven't considered that
`·4· · · ·question carefully and the reason, as a
`·5· · · ·professor of computer science, I'm very
`·6· · · ·loathe to make such a universal statement,
`·7· · · ·is one of the key points in computer science
`·8· · · ·is that the line between data and
`·9· · · ·instructions precisely can be blurry and
`10· · · ·context dependent.· Again, examining a
`11· · · ·particular system I may be able to look at
`12· · · ·it and say, yes, this is not a Downloadable
`13· · · ·in the context of this system.· Trying to
`14· · · ·universalize is kind of dangerous or
`15· · · ·concerning in security scenarios and so when
`16· · · ·you're trying to say can I remove the
`17· · · ·context, I feel uncomfortable doing that.
`18· ·Q.· It's your opinion that the term
`19· · · ·"Downloadable" includes executable files or
`20· · · ·exe files, correct?
`21· ·A.· I think you said executable files or exe
`22· · · ·files and I think you're sort of using
`23· · · ·executable in two different ways.
`24· ·Q.· Sure.· I'm trying to say .exe files.· It's
`
`Page 70
`·1· · · ·your opinion Downloadables would include
`·2· · · ·those, correct?
`·3· · · · · · · · MS. HEDVAT:· Objection, form.
`·4· ·A.· Yes, on many, for instance, Microsoft
`·5· · · ·systems in the exe file, the notes that it's
`·6· · · ·an executable in the sense that it contains
`·7· · · ·computer instructions and could be run on
`·8· · · ·the destination computer.
`·9· ·Q.· Do you know when .exe files first existed?
`10· ·A.· No.· I'd have to look that up.
`11· ·Q.· Did they exist while you were in school?
`12· · · · · · · · MS. HEDVAT:· Objection, form.
`13· ·A.· I honestly cannot recall if they were
`14· · · ·labeled as .exe files.· That was a long time
`15· · · ·ago, both in terms of my memory and back in
`16· · · ·the day of how they worked.· For instance,
`17· · · ·when I was a kid I was working at Microsoft
`18· · · ·when they were developing the early versions
`19· · · ·of Windows, which is what I associate with
`20· · · ·exe files.· I also would have to go back and
`21· · · ·check.· Exe files I assume has some long
`22· · · ·form of history.· Their format or what they
`23· · · ·include might have changed.· I'd have to go
`24· · · ·back to look to see and sort of clarify or
`
`Page 71
`
`·1· · · ·justify the question.
`·2· ·Q.· Some type of executable file existed when
`·3· · · ·you were in school, correct?
`·4· · · · · · · · MS. HEDVAT:· Objection, form.
`·5· ·A.· I'd say some type of -- there were files
`·6· · · ·that would execute and that they would
`·7· · · ·contain computer instructions and run on a
`·8· · · ·computer even when I was a kid.
`·9· ·Q.· And you could download those from the
`10· · · ·internet, correct?
`11· · · · · · · · MS. HEDVAT:· Objection, form.
`12· ·A.· At some point, yes.
`13· ·Q.· At the time you were in school?
`14· ·A.· Yes.· I believe so, that you could download
`15· · · ·programs from the internet.
`16· ·Q.· When did you graduate?
`17· ·A.· Graduated college in '91.
`18· ·Q.· Is an HTML file a Downloadable?
`19· ·A.· Yes.
`20· ·Q.· Even if it doesn't contain script?
`21· ·A.· Again, I would say yes, there are instances
`22· · · ·certainly where it would not contain what
`23· · · ·you would call -- I'm assuming you're
`24· · · ·referring to script tags that would be used
`
`Page 72
`
`·1· · · ·to include script, so I would say yes.
`·2· ·Q.· So an HTML file that doesn't contain any
`·3· · · ·script tags would still be a Downloadable
`·4· · · ·under your understanding?
`·5· ·A.· As -- at the very least, yes, it could be,
`·6· · · ·and as we sometimes discussed with regards
`·7· · · ·to other sort of files, if one wanted to
`·8· · · ·argue there was some limited subset of HTML
`·9· · · ·that might be considered nonexecutable, I
`10· · · ·think I would disagree with that.· It would
`11· · · ·not affect my current infringement analysis
`12· · · ·or I would tend to disagree with that or
`13· · · ·would have to consider that argument, but
`14· · · ·certainly there are certain HTMLs without
`15· · · ·script tags that would still be considered
`16· · · ·executable.
`17· ·Q.· In your infringement theory that's included
`18· · · ·in your declaration, you identified two
`19· · · ·examples of Downloadables, correct?
`20· · · · · · · · MS. HEDVAT:· Objection, form.
`21· ·A.· You could point me.· I can't recall.
`22· ·Q.· It's sort of throughout.· You just -- so my
`23· · · ·understanding is you basically have proposed
`24· · · ·an infringement scenario where a Word file
`
`
`
`Case 3:17-cv-05659-WHA Document 371-15 Filed 02/14/19 Page 7 of 7
`
`Page 197
`
`·1· · · ·same action may be both.
`·2· ·Q.· Do you agree that the sample ID used by Sky
`·3· · · ·ATP as the key for its storage solution does
`·4· · · ·not include the hashes for any dropped files
`·5· · · ·contained in the file?
`·6· · · · · · · · MS. HEDVAT:· Objection, form.
`·7· ·A.· So I think if we're separating out sample ID
`·8· · · ·and clear that that's distinct from
`·9· · · ·Downloadable ID, which I think is what you
`10· · · ·intended from your description, then I think
`11· · · ·I can agree to that statement, but just to
`12· · · ·be clear, those are two distinct things that
`13· · · ·I would not want confused.
`14· · · · · · · · THE WITNESS:· Do you mind if we
`15· · · ·take a break?
`16· · · · · · · · MS. CARSON:· Yes.· We can take a
`17· · · ·break.
`18· · · · · · · · THE VIDEOGRAPHER:· Time now is
`19· · · ·4:09 p.m.· Off the record.
`20· · · · · · · · (Recess.)
`21· · · · · · · · THE VIDEOGRAPHER:· The time now is
`22· · · ·4:17 p.m.· On the record.
`23· · · ·BY MS. CARSON:
`24· ·Q.· So I want to turn to your claim construction
`
`Page 198
`·1· · · ·section on performing a hashing function on
`·2· · · ·the Downloadable and the fetched software
`·3· · · ·components to generate a Downloadable ID.
`·4· ·A.· Uh-hum.
`·5· ·Q.· And I want to first direct you to
`·6· · · ·paragraph 40.
`·7· ·A.· Yes.
`·8· ·Q.· You say "Doctor Rubin's argument also
`·9· · · ·ignores that a POSITA would know that using
`10· · · ·multiple hash functions on multiple
`11· · · ·components to create a single Downloadable
`12· · · ·ID would be both more efficient and
`13· · · ·effective"?
`14· ·A.· Yes.
`15· ·Q.· And then you go on to say "in terms of
`16· · · ·efficiency, it would allow a system to hash
`17· · · ·components that were present if waiting was
`18· · · ·required to obtain other fetched
`19· · · ·components".· Do you see that?
`20· ·A.· Yes.
`21· ·Q.· So I want to confirm that your understanding
`22· · · ·of Claim 1 is that you don't have to
`23· · · ·complete the fetching of the components
`24· · · ·before you perform the hashing function?
`
`Page 199
`
`·1· · · · · · · · MS. HEDVAT:· Objection, form.
`·2· ·A.· So I think to answer that we need to be
`·3· · · ·clear on what we mean by the hashing
`·4· · · ·function.· As I stated, there's the global
`·5· · · ·hashing function that creates a global ID,
`·6· · · ·right?· And when that is based on all the
`·7· · · ·fetch components that will occur after you
`·8· · · ·have fetched all the components as is done,
`·9· · · ·for instance, in the Juniper system to do
`10· · · ·subcomputations, to do parts of the
`11· · · ·computation, it is both reasonable and
`12· · · ·sensible and as a point can be efficient to
`13· · · ·do subcomputations on pieces or subparts,
`14· · · ·whether that's components or otherwise, but
`15· · · ·I guess I feel like you're conflating the
`16· · · ·subhash part with the global hash part in
`17· · · ·that statement.
`18· ·Q.· Let me try a different question.· Is it your
`19· · · ·understanding that you don't have to compete
`20· · · ·the fetching step before you start
`21· · · ·performing the hashing function step?
`22· ·A.· I think that's right.
`23· ·Q.· If you could look at paragraph 41 --
`24· ·A.· Yes.· And by the way, that's true even like
`
`Page 200
`·1· · · ·on single files.· There are hash functions
`·2· · · ·that work on data streams, right, so as data
`·3· · · ·comes in you update your hash as you get
`·4· · · ·more data?· That's like very common and
`·5· · · ·natural, right, so the idea that you have to
`·6· · · ·wait until an entire file arrives before you
`·7· · · ·start hash computation, that's -- that would
`·8· · · ·not be how one of skill in the art would
`·9· · · ·understand hashing in that context.
`10· ·Q.· Okay.· And so that doesn't change your
`11· · · ·answer that you gave before, which is --
`12· ·A.· No.· I was just elaborating.
`13· ·Q.· Okay.· Okay.· So in paragraph 41 you say
`14· · · ·"Doctor Rubin argues that this term should
`15· · · ·be limited to a "single hash value that
`16· · · ·identifies the contents of both the
`17· · · ·Downloadable and the fetched components"".
`18· · · ·Do you see that?
`19· ·A.· Yes.
`20· ·Q.· You say "as such, Doctor Rubin appears to be
`21· · · ·asking to construe the term "Downloadable
`22· · · ·ID""?
`23· ·A.· Yes.
`24· ·Q.· Do you know if any prior courts have
`
`