`Case 5:18—md-02834-BLF Document 410-2 Filed 04/19/19 Page 1 of 23
`
`
`
`
`
`
`
`EXHIBIT 2
`EXHIBIT 2
`
`
`
`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
`28
`
`
`Case 5:18-md-02834-BLF Document 410-2 Filed 04/19/19 Page 2 of 23
`
`
`
`
`
`Michael A. Sherman (SBN 94783)
`
`masherman@stubbsalderton.com
`Jeffrey F. Gersh (SBN 87124)
`
`jgersh@stubbsalderton.com
`Sandeep Seth (SBN 195914)
`
`sseth@stubbsalderton.com
`Wesley W. Monroe (SBN 149211)
`wmonroe@stubbsalderton.com
`Stanley H. Thompson, Jr. (SBN 198825)
`sthompson@stubbsalderton.com
`Viviana Boero Hedrick (SBN 239359)
`vhedrick@stubbsalderton.com
`STUBBS, ALDERTON & MARKILES, LLP
`15260 Ventura Blvd., 20th Floor
`Sherman Oaks, CA 91403
`Telephone:
`(818) 444-4500
`Facsimile:
`(818) 444-4520
`
`Attorneys for Plaintiffs
`[Additional Attorneys listed
`below]
`
`
`UNITED STATES DISTRICT COURT
`
`NORTHERN DISTRICT OF CALIFORNIA
`SAN JOSE DIVISION
`CASE NO.: 5:18-md-02834-BLF
`IN RE PERSONALWEB TECHNOLOGIES,
`
`LLC, ET AL., PATENT LITIGATION
`
`
`FIRST SECOND AMENDED
`
`COMPLAINT
`
`
`DEMAND FOR JURY TRIAL
`_______________________________________
`
`
`
`PERSONALWEB TECHNOLOGIES, LLC, a
`Case No.: 5:18-cv-05625-BLF
`Texas limited liability company, and
`LEVEL 3 COMMUNICATIONS, LLC,
`
`a Delaware limited liability company,
`
`
`Plaintiffs,
`
`v.
`
`ZOOM VIDEO COMMUNICATIONS, INC., a
`Delaware corporation,
`
`
`Defendant.
`
`
`
`
`
`
`FIRST SECOND AMENDED COMPLAINT
`
`
`
`
`
`CASE NO: 5:18-md-02834-BLF
`CASE NO: 5:18-cv-05625-BLF
`
`
`
`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
`28
`
`
`Case 5:18-md-02834-BLF Document 410-2 Filed 04/19/19 Page 3 of 23
`
`
`
`Plaintiff PersonalWeb Technologies, LLC (“Plaintiff” or “PersonalWeb”) files this First
`
`Second Amended Complaint (“Complaint”) for patent infringement against Defendant Zoom Video
`
`Communications, Inc. (“Defendant”). Plaintiff PersonalWeb Technologies, LLC alleges:
`
`
`
`PRELIMINARY STATEMENT
`1.
`PersonalWeb and Level 3 Communications, LLC (“Level 3”) are parties to an
`agreement between Kinetech, Inc. and Digital Island, Inc. dated September 1, 2000 (the “Agreement”).
`Pursuant to the Agreement, PersonalWeb and Level 3 each own a fifty percent (50%) undivided
`interest in and to the patents at issue in this action: U.S. Patent Nos. 6,928,442, 7,802,310, and
`8,099,420 (“Patents-in-Suit”). Level 3 has joined in this Complaint pursuant to its contractual
`obligations under the Agreement, at the request of PersonalWeb.
`2.
`Pursuant to the Agreement, Level 3 has, among other rights, certain defined rights to
`use, practice, license, sublicense and enforce and/or litigate the Patents-in-Suit in connection with a
`particular field of use (“Level 3 Exclusive Field”). Pursuant to the Agreement PersonalWeb has,
`among other rights, certain defined rights to use, practice, license, sublicense, enforce and/or litigate
`the Patents-in-Suit in fields other than the Level 3 Exclusive Field (the “PersonalWeb Patent Field”).
`3.
`All
`infringement allegations, statements describing PersonalWeb, statements
`describing any Defendant (or any Defendant’s products) and any statements made regarding
`jurisdiction and venue are made by PersonalWeb alone, and not by Level 3. PersonalWeb alleges that
`the infringements at issue in this case all occur within, and are limited to, the PersonalWeb Patent
`Field. Accordingly, PersonalWeb has not provided notice to Level 3—under Section 6.4.1 of the
`Agreement or otherwise—that PersonalWeb desires to bring suit in the Level 3 Exclusive Field in its
`own name on its own behalf or that PersonalWeb knows or suspects that Defendant is infringing or
`has infringed any of Level 3’s rights in the patents.
`
`
`
`
`FIRST SECOND AMENDED COMPLAINT
`
`
`
`1
`
`CASE NO: 5:18-md-02834-BLF
`CASE NO: 5:18-cv-05625-BLF
`
`
`
`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
`28
`
`
`Case 5:18-md-02834-BLF Document 410-2 Filed 04/19/19 Page 4 of 23
`
`
`
`THE PARTIES
`
`4.
`Plaintiff PersonalWeb Technologies, LLC is a limited liability company duly organized
`
`and existing under the laws of Texas with its principal place of business at 112 E. Line Street, Suite
`
`
`204, Tyler, TX 75702.
`5.
`Plaintiff Level 3 Communications, LLC is a limited liability company organized under
`the laws of Delaware with its principal place of business at 100 CenturyLink Drive, Monroe,
`Louisiana, 71203.
`6.
`PersonalWeb’s infringement claims asserted in this case are asserted by PersonalWeb
`and all fall outside the Level 3 Exclusive Field. Level 3 is currently not asserting patent infringement
`in this case in the Level 3 Exclusive Field against any Defendant.
`7.
`Defendant Zoom Video Communications, Inc. is, upon information and belief, a
`Delaware corporation having a principal place of business or regular and established place of business
`at 55 Almaden Boulevard, 6th Floor, San Jose, California 95113.
`
`JURISDICTION AND VENUE
`8.
`The court has subject matter jurisdiction pursuant to 28 U.S.C. §§ 1331 and 1338(a)
`because this action arises under the patent laws of the United States, 35 U.S.C. §§ 1 et seq.
`9.
`Venue is proper in this federal district pursuant to 28 U.S.C. §§ 1391(b)–(c) and
`1400(b) because Defendant is incorporated in the State of Delaware and, on information and belief,
`has a regular and established place of business in this District and has committed acts of infringement
`in this District.
`10.
`This court has personal jurisdiction over Defendant because, in addition to the
`allegations in above paragraphs, on information and belief, Defendant is domiciled in this District.
`Further, on information and belief, Defendant purposefully directed activities at residents of
`California, the claims herein arise out of and relate to those activities, and assertion of personal
`jurisdiction over Defendant would be fair.
`
`
`
`
`FIRST SECOND AMENDED COMPLAINT
`
`
`
`2
`
`CASE NO: 5:18-md-02834-BLF
`CASE NO: 5:18-cv-05625-BLF
`
`
`
`Case 5:18-md-02834-BLF Document 410-2 Filed 04/19/19 Page 5 of 23
`
`
`
`PERSONALWEB BACKGROUND
`
`11.
`The Patents-in-Suit cover fundamental aspects of cloud computing, including the
`
`identification of files or data and the efficient retrieval thereof in a manner which reduces bandwidth
`
`
`transmission and storage requirements.
`12.
`The ability to reliably identify and access specific data is essential to any computer
`system or network. On a single computer or within a small network, the task is relatively easy: simply
`name the file, identify it by that name and its stored location on the computer or within the network,
`and access it by name and location. Early operating systems facilitated this approach with standardized
`naming conventions, storage device identifiers, and folder structures.
`13.
`Ronald Lachman and David Farber, the inventors of the Patents-in-Suit, recognized
`that the conventional approach for naming, locating, and accessing data in computer networks could
`not keep pace with ever-expanding, global data processing networks. New distributed storage systems
`use files that are stored across different devices in dispersed geographic locations. These different
`locations could use dissimilar conventions for identifying storage devices and data partitions.
`Likewise, different users could give identical names to different files or parts of files—or unknowingly
`give different names to identical files. No solution existed to ensure that identical file names referred
`to the same data, and conversely, that different file names referred to different data. As a result,
`expanding networks could not only become clogged with duplicate data, they also made locating and
`controlling access to stored data more difficult.
`14.
`Lachman and Farber developed a solution: replacing conventional naming and storing
`conventions with system-wide “substantially unique,” content-based identifiers. Their approach
`assigned substantially unique identifiers to “data items” of any type: “the contents of a file, a portion
`of a file, a page in memory, an object in an object-oriented program, a digital message, a digital
`scanned image, a part of a video or audio signal, or any other entity which can be represented by a
`sequence of bits.” Applied system-wide, this invention would permit any data item to be stored,
`located, managed, synchronized, and accessed using its content-based identifier.
`15.
`To create a substantially unique, content-based identifier, Lachman and Farber turned
`to cryptography. Cryptographic hash functions, including MD4, MD5, and SHA, had been used in
`
`
`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
`28
`
`
`
`FIRST SECOND AMENDED COMPLAINT
`
`
`
`3
`
`CASE NO: 5:18-md-02834-BLF
`CASE NO: 5:18-cv-05625-BLF
`
`
`
`Case 5:18-md-02834-BLF Document 410-2 Filed 04/19/19 Page 6 of 23
`
`
`
`computer systems to verify the integrity of retrieved data—a so-called “checksum.” Lachman and
`
`Farber recognized that these same hash functions could be devoted to a vital new purpose: if a
`
`cryptographic hash function was applied to a sequence of bits (a “data item”), it would produce a
`
`
`substantially unique result value, one that: (1) virtually guarantees a different result value if the data
`item is changed; (2) is computationally difficult to reproduce with a different sequence of bits; and
`(3) cannot be used to recreate the original sequence of bits.
`16.
`These cryptographic hash functions would thus assign any sequence of bits, based on
`content alone, with a substantially unique identifier. Lachman and Farber estimated that the odds of
`these hash functions producing the same identifier for two different sequences of bits (i.e., the
`“probability of collision”) would be about 1 in 2 to the 29th power. Lachman and Farber dubbed their
`content-based identifier a “True Name.”
`17.
`Using a True Name, Lachman and Farber conceived various data structures and
`methods for managing data (each data item correlated with a single True Name) within a network—
`no matter the complexity of the data or the network. These data structures provide a key-map
`organization, allowing for a rapid identification of any particular data item anywhere in a network by
`comparing a True Name for the data item against other True Names for data items already in the
`network. In operation, managing data using True Names allows a user to determine the location of
`any data in a network, determine whether access is authorized, and to selectively provide access to
`specific content not possible using the conventional naming arts.
`18.
`On April 11, 1995, Lachman and Farber filed their patent application, describing these
`and other ways in which content-based “True Names” elevated data-processing systems over
`conventional file-naming systems. The first True Name patent issued on November 2, 1999. The last
`of the Patents-in-Suit has expired, and the allegations herein are directed to the time period before
`expiration of the last of the Patents-in-Suit.
`19.
`PersonalWeb has successfully enforced its intellectual property rights against third
`party infringers, and its enforcement of the Patents-In Suit is ongoing. This enforcement has resulted
`in PersonalWeb obtaining settlements and granting non-exclusive licenses regarding the Patents-in-
`Suit.
`
`
`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
`28
`
`
`
`FIRST SECOND AMENDED COMPLAINT
`
`
`
`4
`
`CASE NO: 5:18-md-02834-BLF
`CASE NO: 5:18-cv-05625-BLF
`
`
`
`Case 5:18-md-02834-BLF Document 410-2 Filed 04/19/19 Page 7 of 23
`
`
`
`GENERAL BACKGROUND
`
`20.
`A webpage is a type of document that is typically retrieved over the World Wide Web,
`
`made viewable and formatted (rendered) by a web browser, and displayed electronically. A “webpage”
`
`
`often refers to what is visible in a browser, but sometimes also refers to a computer file (“webpage
`base file”), usually written in Hypertext Markup Language (“HTML”) or a comparable markup
`language. Such HTML webpage base files typically include text, formatting, and references
`(hyperlinks) to other web content, such as style sheets, scripts, and images that make up part of the
`webpage. Web content referenced in an HTML or similar file are also called “asset files” herein. The
`web browser coordinates the retrieval of the various asset files of a webpage and renders the webpage
`for display from the webpage base file and the asset files referenced in the webpage base file or
`referenced in other asset files.
`21.
`On the World Wide Web, hyperlinks generally include Uniform Resource Identifiers
`(“URIs”), which each typically include an address of a server (“host”) from which the asset file is to
`be retrieved (e.g., “www.website.com”), a “path” to the location of that asset file on the host server
`(e.g., “/directory/”), and a filename (e.g., “filename.ext”).
`22.
`On the Internet, a web browser typically retrieves a webpage base file from a remote
`web server and retrieves referenced asset files from the same or different servers. The web browser
`retrieves a webpage base file or an asset file by making a GET “request” to a web server using the
`Hypertext Transfer Protocol (“HTTP”), an industry standard. The web server may respond to such an
`HTTP request with a HTTP “response” that includes the requested web content and may include other
`information or instructions.
`23.
`A static webpage is delivered exactly as stored, as web content in the web server’s file
`system or memory. In contrast, a dynamic webpage is generated by a web server application, usually
`driven by server-side software, upon receipt of a request from a browser (user). For example, a picture
`of a building might be delivered as static content (a picture) whereas the latest traffic conditions may
`be delivered dynamically based on real time traffic information.
`24.
`The speed of a browser retrieving webpage base files and incorporated asset files can
`be increased by the browser storing previously retrieved webpage base files and asset files in a browser
`
`
`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
`28
`
`
`
`FIRST SECOND AMENDED COMPLAINT
`
`
`
`5
`
`CASE NO: 5:18-md-02834-BLF
`CASE NO: 5:18-cv-05625-BLF
`
`
`
`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
`28
`
`
`Case 5:18-md-02834-BLF Document 410-2 Filed 04/19/19 Page 8 of 23
`
`
`
`“cache” on the computer running the browser. If a browser’s user later requests a previously retrieved
`
`webpage base file or requests a webpage that includes an asset file previously used by the browser in
`
`rendering the same or a different webpage (for example, by reloading a webpage or visiting the same
`
`
`webpage again), the browser may use the cached webpage base file or asset file rather than having to
`download the same file repeatedly over the Internet again.
`25.
`Two computers communicating over the Internet usually are not directly connected to
`each other but rather interact via chains of network appliances and other computers (e.g., “switches”
`and “intermediate” servers). Many intermediate servers have caches similar to and complementing
`the browser cache that store webpage base files and assets that pass through that intermediate server.
`If a browser or server requests a file from the intermediate server that is present in that intermediate
`server’s cache, the intermediate server can use the content in its cache to respond to the request rather
`than send the request upstream towards the web server from which the file initially originated (also
`called the “origin server”).
`26.
`Responses to HTTP requests may include header elements (control elements) and a
`body (the “object” that was requested). Under HTTP, web servers can include a “cache-control”
`header with a response that includes a webpage or asset file. A “cache-control” header includes one
`or more directives that instruct browsers and intermediate server caches (“intermediate caches”) as to
`whether and for how long the file (object) included in the response may be cached or under what
`circumstances and under what conditions the cached content may be used. HTTP also provides for
`including other headers in responses that provide similar types of instructions to browsers and
`intermediate caches. Collectively, these other headers and directives in a “cache-control” header are
`referred to herein as “cache-control headers.”
`27.
`Given that webpage content changes, sometimes rather quickly and regularly, a
`problem that website owners face is effectively instructing a browser that is re-rendering a previously
`cached webpage that one or more of its cached files for that webpage are no longer the correct and
`authorized content (the content of those files has changed) and similarly reauthorizing the use of those
`cached files whose content has not changed.
`
`
`
`
`FIRST SECOND AMENDED COMPLAINT
`
`
`
`6
`
`CASE NO: 5:18-md-02834-BLF
`CASE NO: 5:18-cv-05625-BLF
`
`
`
`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
`28
`
`
`Case 5:18-md-02834-BLF Document 410-2 Filed 04/19/19 Page 9 of 23
`
`
`
`28.
`On one hand, website owners want to encourage the browsers that render their web
`
`pages to use cached files thereby reducing the number of requests for these files that are being made
`
`to their webpage servers. Therefore, they frequently will set cache-control headers that authorize the
`
`
`browser to cache their webpage base files and asset files so the files are on hand when the browser
`needs to render that webpage again. On the other hand, website owners want the browsers to use the
`latest authorized files so that their users do not see the wrong content when viewing their webpage.
`
`DEFENDANT’S BACKGROUND
`On information and belief, Defendant has operated a website located at zoom.us, and
`29.
`has done so since before expiration of the last to expire of the Patents-in-Suit, which has operated to
`provide authorized webpage content to its users in the manner herein described.1
`30.
`On information and belief, Defendant’s web servers utilized a system of notifications
`and authorizations to control the distribution of content, e.g., what webpage content may be served
`from web servers and intermediate caches and what cached webpage content a browser is re-authorized
`to use to render Defendant’s webpage(s).
`31.
`On information and belief, Defendant’s system and its associated method of providing
`webpage content used “conditional” HTTP GET requests with If-None-Match headers and associated
`content-based ETag values for various asset files required to render various webpages of the
`Defendant.
`32.
`On information and belief, Defendant’s system and associated method used these
`ETags to instruct both the intermediate cache servers and the endpoint caches at browsers to verify
`whether they were still authorized to reuse the previously cached webpage base files of Defendant and
`to instruct them to obtain newly authorized content in rendering Defendant’s webpage when that
`content had changed. In other words, whether the previously cached content was still considered valid
`for use by the Defendant website operator.
`
`
`1 While the complaint is sometimes written in the present or present perfect tense, all specific
`allegations are directed to the system’s operations and the method’s performance in the relevant time
`
`period.
`
`FIRST SECOND AMENDED COMPLAINT
`
`
`
`CASE NO: 5:18-md-02834-BLF
`CASE NO: 5:18-cv-05625-BLF
`
`7
`
`
`
`Case 5:18-md-02834-BLF Document 410-2 Filed 04/19/19 Page 10 of 23
`
`
`
`33.
`On information and belief, Defendant thereby reduced the bandwidth and computation
`
`required by its origin servers and any intermediate cache servers to field user requests to render
`
`Defendant’s webpages as those servers only need to serve files whose content has changed. On
`
`
`information and belief, this has allowed for the efficient update of cached information only when such
`content has changed, thereby reducing transaction overhead and bandwidth and allowing the
`authorized content to be served from the nearest cache.
`34. More particularly, on information and belief, each of Defendant’s webpages included
`a webpage base file (e.g., a main or initial HTML file) and one or more asset files referenced in the
`webpage base file (or referenced in other asset files that contained references to other asset files). On
`information and belief, the references in the webpage base file to the asset files needed to render the
`webpage were typically Uniform Resource Identifiers (“URIs”), which each typically included a
`filename, the address of a host server from which the asset file could be retrieved, and a “path” to the
`location of that asset file on that server.
`35.
`On information and belief, for at least one of the asset files (“CBI ETag asset files”),
`the asset file comprised a sequence of bits and an associated ETag value was generated by Defendant
`by applying a hash function to the sequence of bits; wherein any two CBI ETag asset files comprising
`identical sequences of bits had identical associated ETag values. Thus, on information and belief,
`when a CBI ETag asset file’s content was changed a new associated ETag value was generated by
`Defendant. On information and belief, Defendant caused the origin server for each CBI ETag asset
`file to serve such CBI ETag asset file with its associated EtagETag value in response to HTTP GET
`requests for the CBI ETag asset file.
`36.
`On information and belief, Defendant contracted with Amazon to use Amazon’s S3
`system to store and serve at least some of Defendant’s CBI ETag files (“S3 asset files”) on its behalf.
`On information and belief, once Defendant’s S3 asset files were compiled and are complete, Defendant
`uploaded them to an Amazon S3 server as objects. On information and belief, such objects comprised
`a sequence of bits and, upon upload, an associated ETag value was generated by the S3 system on
`behalf of Defendant by applying a hash function to the sequence of bits, wherein any two S3 asset
`files comprising identical sequences of bits had identical associated ETag values. On information and
`
`
`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
`28
`
`
`
`FIRST SECOND AMENDED COMPLAINT
`
`
`
`8
`
`CASE NO: 5:18-md-02834-BLF
`CASE NO: 5:18-cv-05625-BLF
`
`
`
`Case 5:18-md-02834-BLF Document 410-2 Filed 04/19/19 Page 11 of 23
`
`
`
`belief, in this way, Defendant generated the associated ETag values for its CBI ETag asset files that
`
`were S3 asset files.
`
`36.37. On information and belief, , Defendant also contracted with Amazon to use Amazon’s
`
`
`CloudFront system to act as the origin server on its behalf instead of S3 for at least some of Defendant’s
`S3 asset files (“CloudFront served asset files”). On information and belief, a CloudFront the S3 server,
`for each CloudFront served S3 asset file, served the CloudFront served S3 asset file with the its
`associated ETag value in response to HTTP GET requests for the CloudFront served S3 asset file.
`37.38. On information and belief, when an intermediate cache server or a browser requested
`a webpage from the Defendant for the first time, it sent an HTTP GET request with the webpage’s
`URI and Defendant’s origin server or an upstream cache server responded by sending an HTTP 200
`(OK) response message containing the webpage base file. On information and belief, a browser then
`sent individual HTTP GET requests, each with an asset file’s URI that was referenced in the webpage
`base file, and the asset files’ origin servers or intermediate cache servers responded by sending
`individual HTTP 200 responses containing the requested asset files, along with, if available, their
`respective associated ETags. On information and belief, in the case of CloudFront served asset files,
`the asset files’ origin server was a CloudFront server. On information and belief, at least some of
`Defendant webpage base files included references to CloudFront served asset files. On information
`and belief, upon receipt of the HTTP 200 responses, the intermediate cache server or browser cached
`the webpage base file and asset files with their associated URI and associated ETag values and the
`browser used them in rendering the requested web page of the Defendant. On information and belief,
`the origin servers, intermediate cache servers, and browser caches were caused to maintain
`databases/tables which mapped the URIs of webpage base files and asset files to their respective
`responses and, if applicable, associated cache-control headers and ETags.
`38.39. On information and belief, by responding to an HTTP GET request for a given webpage
`by transmitting content of an asset file with an associated ETag, Defendant instructed the browser
`cache and all intermediate cache servers, to use an HTTP conditional GET request the next time that
`asset file is requested. More specifically, on information and belief, the browser or intermediate cache
`is instructed to include the ETag in the HTTP conditional GET request with an “If-None-Match”
`
`
`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
`28
`
`
`
`FIRST SECOND AMENDED COMPLAINT
`
`
`
`9
`
`CASE NO: 5:18-md-02834-BLF
`CASE NO: 5:18-cv-05625-BLF
`
`
`
`Case 5:18-md-02834-BLF Document 410-2 Filed 04/19/19 Page 12 of 23
`
`
`
`header to re-verify that they are still authorized to serve or use that content or determine that they are
`
`no longer authorized to use that content and therefore must use new content.
`
`39.40. On information and belief, Defendant did this, for example, by causing cache-control
`
`
`headers to be included in HTTP responses containing its asset files. On information and belief,
`Defendant benefits from using the ETags to control the distribution of its webpage content by
`communicating to a downstream cache and to a browser which of Defendant’s cached webpage base
`files it is reauthorized to serve/use and what newly authorized files it must first obtain in
`serving/rendering Defendant’s webpages.
`40.41. More particularly, on information and belief, when a browser again requested the
`Defendant’s webpage, the browser either used a cached copy, if allowed by the cache-control headers,
`or retrieved a new copy of the webpage base file for Defendant’s webpage. Similarly, on information
`and belief, for asset files referenced in the new or cached webpage base file, the browser either used a
`cached copy, if allowed by the cache-control headers, or retrieved a new copy of the asset files for
`Defendant’s webpage.
`41.42. On information and belief, for an asset file stored in the browser’s cache with an ETag,
`and based on the cache-control headers received in the original response, the browser sent a conditional
`GET request with an If-None-Match header using the associated ETag value and the URI for the asset
`file so as to be notified whether the browser still had Defendant’s authority to render the webpage with
`its locally cached asset file. In other words, whether the cached content was still valid for use in
`rendering Defendant’s webpage.
`42.43. On information and belief, under most circumstances, a responding intermediate cache
`server having content cached for the URI in the conditional GET request and having an ETag for that
`URI responded to the request by determining whether it had the same associated ETag value for that
`URI. If it had no ETag value for that URI, on information and belief, the request was passed up to an
`upstream intermediate cache server capable of responding or, if none, to the URI’s origin server, which
`responded to the request. On information and belief, if the intermediate cache server did not have
`content cached for the URI in the conditional GET request, the request was similarly passed up to an
`upstream intermediate cache server capable of responding or, if none, to the URI’s origin server.
`
`
`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
`28
`
`
`
`FIRST SECOND AMENDED COMPLAINT
`
`
`
`10
`
`CASE NO: 5:18-md-02834-BLF
`CASE NO: 5:18-cv-05625-BLF
`
`
`
`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
`28
`
`
`Case 5:18-md-02834-BLF Document 410-2 Filed 04/19/19 Page 13 of 23
`
`
`
`43.44. On information and belief, if the responding server had the webpage content for that
`
`URI and there was a match between the ETag it received in the request with the ETag it currently had
`
`associated for that URI, it sent back an HTTP 304 (Not Modified) response message; this message
`
`
`notifying the browser that the same webpage content was present at the responding server and that the
`browser was still authorized to use that previously cached asset file to render the webpage. On
`information and belief, upon receipt of the HTTP 304 response, the browser accessed the locally
`cached asset file in rendering the webpage.
`44.45. On information and belief, if the asset file’s associated ETag sent by the browser in the
`conditional GET If-None-Match request did not match the associated ETag maintained at the
`responding server (or other intermediate cache servers further upstream or the origin server) for that
`URI, the responding server sent back an HTTP 200 response along with the new asset file and its new
`ETag value. The HTTP 200 response indicated to the browser that it was not authorized to use (or
`serve, in the case of an intermediate cache server receiving the HTTP 200 response) the previously
`cached asset file. In response to receiving the HTTP 200 response, the browser (or intermediate cache
`server) was instructed to update its respective cache with the new asset file and associated ETag. The
`browser subsequently used the new asset file to render the webpage.
`45.46. Exhibit 1 to the complaint lists a specific example of a file that was, on information and
`belief, served by or on behalf of Defendant during the relevant time period. The example in Exhibit
`1 is an asset file served by CloudFront S3 with a content-based ETag generated by S3 for that asset
`file.
`
`46.47. On information and belief, in this manner, Defendant used ETag values based on the
`asset files’ content to control the behavior of downstream intermediate cache servers and browser
`caches to assure that they only accessed and used Defendant’s latest authorized webpage content to
`serve or to render its webpages.
`
`
`
`
`FIRST SECOND AMENDED COMPLAINT
`
`
`
`11
`
`CASE NO: 5:18-md-02834-BLF
`CASE NO: 5:18-cv-05625-BLF
`
`
`
`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
`28
`
`
`Case 5:18-md-02834-BLF Document 410-2 Filed 04/19/19 Page 14 of 23
`
`
`
`
`
`
`herein.
`
`FIRST CLAIM FOR RELIEF
`INFRINGEMENT OF U.S. PATENT NO. 6,928,442
`47.48. PersonalWeb repeats and realleges paragraphs 1–4746, as if the same were fully stated
`
`
`48.49. On August 9, 2005, United States Patent No. 6,928,442 (the “’442 patent”) was duly
`and legally issued for an invention entitled “Enforcement and Policing of Licensed Content Using
`Content-Based Identifiers.” PersonalWeb has an ownership interest in the ’442 patent by assignment,
`including the exclusive right to enforce the ’442 patent within the PersonalWeb Patent Field, and
`continues to hold that ownership interest in the ’442 patent.
`49.50. Defendant has infringed at least claims 10 and 11 of the ’442 patent by its manufacture,
`use, sale, importation, and/or offer for sale of products or services, and/or controlling the distribution
`of its webpage content in the manner described herein. Defendant’s infringement is literal and/or
`under the doctrine of equivalents and Defendant is liable for its infringement of the ’442 patent
`pursuant to 35 U.S.C. § 271.
`50.51. For example, claim 10 covers “a method, in a system in which a plurality of files are
`distributed across a plurality of computers.” On information and belief, Defendant has used a system
`of notifications and authorizations to distribute a plurality of files, e.g., D