UNITED STATES DISTRICT COURT FOR THE DISTRICT OF COLUMBIA
July 15, 2009
DOW JONES & COMPANY, INC., PLAINTIFF,
ABLAISE LTD., ET AL., DEFENDANTS.
DOW JONES REUTERS BUSINESS INTERACTIVE, LLC, D/B/A FACTIVA PLAINTIFF,
ABLAISE LTD., ET AL., DEFENDANTS.
The opinion of the court was delivered by: James Robertson United States District Judge
Ablaise is the owner of U.S. Patent No. 6,961,737 ('737 patent) and U.S. Patent No. 6,295,530 ('530 patent). In 2006, Ablaise accused Dow Jones of infringing the patents and offered a licensing agreement. Dow Jones declined the offer and sued for a declaratory judgment that the patents were invalid and not infringed. Ablaise counterclaimed for infringement of both patents.
I issued a Markman order in July 2007. Dkt. 28. In October 2008, I denied Ablaise's motion to dismiss the '530 patent, finding that Ablaise's offer of a covenant not to sue Dow Jones did not oust me of subject matter jurisdiction over Dow Jones' claim that the patent was invalid. See Dkt. 97, at 5.
Now before me is Dow Jones' motion for summary judgment of invalidity as to all asserted claims: Claims 1, 3, 4, and 6 of the '737 patent and Claims 1-3 of the '530 patent. The motion will be granted.
The '737 and '530 patents describe methods for using a computer server to send personalized content and format over the World Wide Web in the form of HTML pages that are generated dynamically. Both patents claim a priority date of May 15, 1995. Statement of Undisputed Facts [hereinafter "SUF"] ¶ 24. At that time, web browsing was fairly common, and web sites were capable of generating personalized content "dynamically" -- that is, without requiring manual alterations to the source code. Id. ¶ 7. But, in Ablaise's telling, web sites, unlike non-Web programs such as Microsoft Access and SQL, id. ¶¶ 14-15, could not display personalized content in a personalized format. The patents-at-issue provided methods for addressing this limitation.
The patented methods involve the use of HTTP, HTML, and CGI. HTTP is an Internet protocol that allows a user (known as a "client") to communicate with a server and request that the server send specific content. HTML is a rendering language that allows the server to define the content and format of a web page through the use of "tags," like "center" to center text and images, or "p" to move text to a new paragraph. CGI is a program that generates web pages dynamically, or "on the fly." Each of these technologies was well known to web developers in May 1995. Id. ¶¶ 1-6.
Although Dow Jones challenges the validity of Claims 1, 3, 4, and 6 of the '737 patent, and Claims 1-3 of the '530 patent, the focus of the parties' motions is exclusively on the validity of the first claim in each of the patents. They agree that, if Claim 1 of either patent is invalid, the remaining asserted claims of that patent are invalid as well.
Claim 1 of the '737 patent describes: A method for serving pages of viewable data to browsing devices connected to a network, wherein a page of said viewable data comprises content data defining text and/or graphics and formatting data which specifies locations of said text and/or graphics with a page, and said viewable data is displayed at a browsing device such that locations of said text and/or graphics depend on said formatting data, said method comprising: identifying requests from browsing devices that define a request for specified content data; storing content data; storing executable functions; maintaining a user database comprising information relating to user preferences; and in response to identifying a request for specified content data and a user identifier: (a) reading user preference information from said user database in response to a received user identifier; (b) selecting stored content data in dependence upon the content data specified in a received request; (c) receiving format identifiers identifying the type of formatting required; (d) selecting a set of stored functions in dependence upon a received format identifier and said read user information; and (e) executing said set of functions to generate viewable data comprising said selected content data and formatting data.
Dkt. 95, Ex. 1, at 19:65-20:25. Claim 4 teaches a "serving device" that performs the method described in Claim 1. Claim 3 is identical to Claim 1 except that it requires that the "viewable data" be "HTML data" and that the "formatting data" comprise "HTML tags." Claim 6 adds the same "serving device" requirement to Claim 3 that Claim 4 adds to Claim 1. Id. at 20:31-21:7.
Claim 1 of the '530 patent describes: A method of serving output signals from a serving device to a plurality of browsing devices connected to a network, wherein said output signals represent commands executable by a browsing device so as to display viewable data in accordance with a specified page format, said method comprising steps of: identifying requests from browsing devices that define a request for specified viewable data, said request including formatting type identification data; maintaining a plurality of formatting types of data defining respectively corresponding predetermined formats for portions of said viewable data; storing content data representing said viewable data; selecting a specific one of said types of formatting data in response to said formatting type identification data; processing said content data and said formatting types of data so as to combine said selected part of said content data with said specific one of said types of formatting data, and for outputting processed viewable data with executable instructions; and supplying output signals to the requesting browser device derived from said output processed data, in which said output signals represent commands executable by a browsing device so as to display said viewable data in accordance with a first specified page format when a first type of formatting data is selected and in a second specified page format when a second type of formatting data is selected.
Dkt. 95, Ex. 2, at 19:55-20:29. Claim 2 requires that the "content data" include "graphics data." Claim 3 requires that a "serving device" perform the method described in Claim 1. Id. at 20:30-21:24.
The patents differ from one another in two relevant respects. First, the '737 patent describes a server that is capable of more flexible formatting than is the '530 patent. If, for example, a user expresses a preference for certain content to appear higher on the web page, the server can generate a web page that satisfies those preferences, while the '530 patent limits a user's formatting options to certain templates. The '530 patent user can choose between available templates, but cannot express more particular formatting preferences.
Second, only the '737 patent describes a server that stores a user's preferences in a database. Once a user expresses a preference for particular content to appear in a particular location, the content will appear in that location whenever the user visits the website until or unless the user expresses a different preference. The '530 patent describes a server that is only capable of accommodating a user's current request for a particular template. If a user requests a specific template, the page will appear in that template at that time, but the user will have to re-request a template the next time he visits the page.
A. Legal Standards
On this motion for summary judgment, all reasonable inferences are drawn in favor of Ablaise, the non-moving party. Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 248 (1986).
Dow Jones contends that the patent claims are either invalid as anticipated, or, if not anticipated, invalid as obvious. Invalidity must be proven "with facts supported by clear and convincing evidence." Adenta GMBH v. Orthoarm, Inc., 501 F.3d 1364, 1371 (Fed. Cir. 2007) (quoting SSIH Equip. S.A. v. U.S. Int'l Trade Comm'n, 718 F.2d 365, 375 (Fed. Cir. 1983)). To prevail on its motion, Dow Jones must first identify relevant prior art. In this case, art is only prior art if patented, used, or published before the priority date of the challenged patents. See 35 U.S.C. § 102(a).*fn1 The art was "used" if there is evidence of "any use of the [art] by a person other than the inventor who is under no limitation, restriction or obligation of secrecy to the inventor." Adenta, 501 F.3d at 1371. The art was "published" if it was "made available to the extent that persons interested and ordinarily skilled in the subject matter or art exercising reasonable diligence [could] locate it and recognize and comprehend therefrom the essentials of the claimed invention without need of further research or experimentation." Bruckelmyer v. Ground Heaters, Inc., 445 F.3d 1374, 1378 (Fed. Cir. 2006).
Dow Jones then must prove that the prior art anticipates the patents. Under 35 U.S.C. § 102, a patent is anticipated if "all of [its] elements and limitations . . . [are] expressly or inherently described in a single prior art reference." Elan Pharmaceuticals, Inc. v. Mayo Found. for Med. Educ. & Research, 304 F.3d 1221, 1227 (Fed. Cir. 2002). The reference "must not only disclose all the elements of the claim within the four corners of the document, but must also disclose those elements 'arranged as in the claim.'" Net MoneyIN, Inc. v. VeriSign, Inc., 545 F.3d 1359, 1369 (Fed. Cir. 2008) (quoting Connell v. Sears, Roebuck & Co., 722 F.2d 1542, 1548 (Fed. Cir. 1983)).
If Dow Jones fails to prove anticipation, it can still prevail if it proves that the patents are obvious. Under 35 U.S.C. § 103(a), Dow Jones must show that "the differences between the subject matter sought to be patented and the prior art are such that the subject matter as a whole would have been obvious at the time the invention was made to a person having ordinary skill in the art to which said subject matter pertains." This § 103 analysis requires an assessment of the "interrelated teachings of [the prior art]; the effects of demands known to the design community or present in the marketplace; and the background knowledge possessed by a person having ordinary skill in the art." KSR Int'l. Co. v. Teleflex, Inc., 127 S.Ct. 1727, 1740 (2007).
B. The '737 patent
Fishwrap was an online newspaper developed at M.I.T. that used CGI programs to generate HTML-formatted pages dynamically. SUF ¶ 35. In its initial form, it allowed users to select the sections and topics they wanted the web page to display. Id. ¶ 34. Later, developers added a "self- organization feature," which recorded a user's reading habits in a database, fed that information into an algorithm, and used the algorithm's output to re-format the newspaper to reflect the user's implied preferences. For instance, if a user viewed sports stories first, or more frequently, then sports stories would eventually appear higher on the page.
a. Prior Art
Dow Jones asserts that Fishwrap is prior art because it was in use before May 15, 1995. Ablaise concedes that versions of Fishwrap were in use before that date, and that some of those versions "self-organized the order of topics generated by the newspaper," Dkt. 99, at 15. But Ablaise maintains that Dow Jones has not shown by clear and convincing evidence that the version that it claims anticipates the '737 patent was in use before the priority date of the patent.
In the opinion of Dow Jones' expert, Dr. Aver Bestavros, it was. Dr. Bestavros bases his opinion on three documents: the source code, an article by Dr. Pascal Chesnais (the inventor of Fishwrap), and an M.I.T. student thesis by Douglas B. Koen. See Bestavros Expert Report, Dkt. 95, Ex. 7, ¶¶ 81-84. The relevant portions of the source code were last modified as late as 1997, see Decl. of Dr. Pascal Chesnais, Dkt. 91, Exs. 40-47, and the Chesnais article was published in June 1995, see id., Ex. 32, so neither document pre-dates the priority date of the challenged patent. And while Koen's thesis was submitted to the M.I.T. Thesis Committee before the priority date, it does not explicitly analyze the source code Dow Jones has provided. See id., Ex. 15. Thus, to establish that its invalidity case is based on prior art, Dow Jones must show that Dr. Bestavros analyzed a version of Fishwrap that existed before May 15, 1995.
Dr. Chesnais states that "[a]ll the functionality . . . relied upon by Dr. Bestavros in his opinion of invalidity was present on May 15, 1995 as Fishwrap Release Four functionality," id. ¶ 50; that the post-May 1995 modifications to the source code were "minor and completely unrelated to the functionality upon which Dr. Bestavros bases his opinion," id. ¶ 53; and that the Koen thesis -- submitted in May 1994 -- "describes all of the features of Fishwrap upon which Dr. Bestavros relied," id. ¶ 49.
Citing Juicy Whip v. Orange Bang, 292 F.3d 728, 743 (Fed. Cir. 2002), and Finnigan v. Int'l Trade Comm'n, 180 F.3d 1354, 1365-68 (Fed. Cir. 1999), Ablaise contends that Dr. Chesnais' statements are "uncorroborated," and that, standing alone, they are not clear and convincing evidence that all of the relevant elements of Release Four were in use before May 1995. However, unlike the cases that Ablaise cites, in which the witnesses offered no contemporaneous documents to support their claims, Dr. Chesnais has offered two contemporaneous documents as corroboration: the Koen thesis and the source code. His claim that the Koen thesis describes all of the features that Dr. Bestavros relies upon in his analysis is verifiable by comparing the thesis and the expert report, and his claim that the post-May 1995 modifications to the source code were immaterial is verifiable by examining those modifications, which are marked in the code.
But, Ablaise goes on to argue, those supposedly corroborating documents do not support Chesnais' statements. Ablaise's expert, Geoff Mulligan, points out that the Koen thesis could not have analyzed Fishwrap Release Four because, according to Dr. Chesnais himself, Release Four was only developed in 1995, see Dkt. 91, ¶ 47, and Koen submitted his thesis in 1994. Decl. of Geoff Mulligan, Dkt. 99, ¶ 10.
Mr. Mulligan also argues that there could have been post-May 1995 modifications to the source code that were simply left unmarked. Id. ¶ 6.
Ultimately, however, the source code and the Koen thesis provide a sufficient basis for understanding the state of Fishwrap before the priority date of the '737 patent. Even if the thesis did not examine what Dr. Chesnais refers to as Fishwrap Release Four, it necessarily analyzed a version of Fishwrap that existed before May 15, 1995. And Mulligan's argument about unmarked modifications in the source code is pure speculation. In the absence of any evidence that there are unmarked modifications to the code, or that the marked post-May 1995 modifications affected the relevant functionality of Fishwrap, I find that the source code, the Koen thesis, and Dr. Chesnais' statements provide clear and convincing evidence that Dr. Bestavros' analysis is based on a version of Fishwrap that is prior art.
Ablaise concedes that Fishwrap practiced every limitation of Claim 1 except its use of a "format identifier." Claim 1 describes a server that, "in response to identifying a request for specified content data and a user identifier," "receiv[es] format identifiers identifying the type of formatting required." In my Markman order, I defined the words "type of formatting" to mean "a layout or presentation of text and/or graphics on a page." Dkt. 28, at 7. Thus, to anticipate Claim 1, Fishwrap must have received a "format identifier" that identified the layout or presentation of text and/or graphics on a page.
On this point, Dow Jones does not sustain its burden of proof. Dr. Bestavros contends that Fishwrap's recorded observations of a user's preferences are "'format identifiers' because they dictate the layout of the newspaper." See Dkt. 95, Ex. 7, ¶ 84. Dow Jones reiterates that claim in its motion, stating that the "stored observations made by Fishwrap unquestionably perform the two functions that comprise the role of the 'format identifier' [because] [t]he Fishwrap system . . . stored information about an individual user's viewing habits on Fishwrap, which it then used to identify the order (layout) in which that user's articles will appear." Dkt. 95, at 27-28; see also id., at 29 ("there is no dispute that Fishwrap's user information performed the functions of a 'format identifier.'").
But, as Dr. Chesnais himself explains, user preferences were not the only factor that determined page layout: "the reorder function [the algorithm] . . . order of sections and topics based in part on relative frequencies and sequences of a user's browsing to the various sections and topics." Dkt. 91, ¶ 55 (emphasis added). Ablaise's expert, Mr. Mulligan, notes that, in addition to user preferences, Fishwrap's algorithm also considered the "newness" of the content -- how recently it had become available to be viewed. Dkt. 99, ¶ 13. As he explained in his deposition, "newness was almost the [algorithm's] top priority," so two users with very different viewing histories could end up receiving content in the same order. Dkt. 95, Ex. 3, at 100:23-101:13. We are left with a genuine dispute of fact as to whether Fishwrap's "format identifier" -- its record of a user's viewing preferences --identified "the type of formatting required."
Dow Jones argues that Fishwrap still anticipates Claim 1 even if recorded observations were not the only factor in the re-ordering process. But I see no other way to read Claim 1's limitation that the format identifier "identify the type of formatting required." The claim language requires that the format identifier specify -- or "dictate," as Dr. Bestavros puts it -- the layout of content on the page. If Mr. Mulligan is correct, the recorded observations of user behavior do not play that role in Fishwrap because they can be counteracted or even overridden by the "newness" of the content.
Dow Jones is right to note that in Mr. Mulligan's analysis Fishwrap is arguably a more advanced version of the server described in Claim 1: rather than re-formatting solely in response to a user's express preferences, it re-formatted based on a user's implied preferences and the "newness" of the available content. But Dow Jones does not get credit, at least towards its anticipation case, for discovering prior art that is more complex than the patented invention. To anticipate the patented invention, the prior art must "disclose all the elements of the claim within the four corners of the document," Net MoneyIN, Inc., 545 F.3d at 1369, and there is a legitimate factual dispute over whether Fishwrap meets that requirement.
Dow Jones can still prevail if it can show by clear and convincing evidence that Fishwrap -- in combination with other prior art references -- renders Claim 1 obvious. "The combination of familiar elements according to known methods is likely to be obvious when it does no more than yield predictable results." KSR, 127 S.Ct. at 1739. But courts "must still be careful not to allow hindsight reconstruction of references to reach the claimed invention without any explanation as to how or why the references would be combined to produce the claimed invention." Innogenetics, N.V. v. Abbott Laboratories, 512 F.3d 1363, 1374 n.3 (Fed. Cir. 2008).
At minimum, Dow Jones must show that a prior art reference disclosed the use of a "format identifier"; that one of ordinary skill in the art would have been aware of that reference; that a person of ordinary skill in the art would be capable of combining the "format identifier" element with Fishwrap; and that a person of ordinary skill in the art would have an "apparent reason to combine the known elements in the fashion claimed by the patent at issue." KSR, 127 S.Ct. at 1740.
As an initial matter, the parties disagree on the experience level and the area of expertise of a person of ordinary skill in the art. Ablaise argues that someone with as little as one year of experience in web page generation would have ordinary skill, see Decl. of Sharon Davis, Ex. 10, ¶¶ 7-9, while Dow Jones contends that it would be someone more knowledgeable in the areas of networked information systems and web page development, see id., Ex. 7, ¶ 80. But more importantly, in his expert report, Dr. Bestavros offers only one paragraph of rather conclusory analysis on why Fishwrap would render Claim 1 obvious:
Performing the extra step of analyzing the user's online behavior to determine the user's preferences to be stored in the database is a more difficult technical process than a system that would allow a user to simply select the order for displaying material directly. A person of ordinary skill in the art would undoubtedly understand that one could very easily allow the user to select the order himself with a simple form allowing the user to select the order in place of Fishwrap's use of behavior analysis. The use of such forms to collect information about a particular user was well known to persons of skill in the art before May 15, 1995. By using a simple form to obtain formatting preferences for the user and recording that information in the database in the place of the preferences recorded by the Fishwrap algorithm, one would achieve the same result (ordering sections based on the user's preferences) more easily.
Dkt. 95, Ex. 7, ¶ 94 (emphasis in original).
That paragraph leaves much unanswered: Which specific prior art references disclose the use of forms to record a user's express preferences? Would those forms meet the '737 patent's requirement of a "format identifier"? How would one of ordinary skill in the art integrate the "format identifier" into Fishwrap's source code? Why would one want to make such a change to Fishwrap's structure?
Dow Jones never addresses how one would integrate a "format identifier" into Fishwrap, but it does attempt to provide answers to the other questions. It contends that the HTGrep and Thunderstone references (which it discusses in greater detail in its invalidity challenge to the '530 patent) disclose the use of "format identifiers." Dkt. 95, at 33. And it submits that a person of ordinary skill in the art would have been motivated to integrate a "format identifier" into Fishwrap because it would have been a "simpler" way of providing users with personalized content. Id.
There is no clear and convincing evidence to support either of these claims. Dr. Bestavros discusses the HTGrep and Thunderstone references in relation to the '530 patent, not the '737 patent, so he does not offer an opinion on whether those references meet the "format identifier" limitation. See id., Ex. 7, ¶¶ 130, 147. And while Dow Jones claims that one would be motivated to add a "format identifier" to Fishwrap to make it simpler, the Koen thesis argues that such a change would actually make things more complicated:
Another user suggested that the problem of paper structuring could be satisfactorily solved by providing the user with a mechanism for moving the topics and sections around manually, and in this way he would be able to select the paper structure that appealed to him. While this solution might be equitable for technically inclined users, the object of this research is to increase automation and responsiveness to the reader's wishes without the need for explicit intervention on his part.
Dkt. 91, Ex. 15, at 26; see DePuy Spine, Inc. v. Medtronic Sofamor Danek, Inc., 567 F.3d 1314, 1327-28 (Fed. Cir. 2009) (finding that a combination of prior art elements did not render the challenged patent obvious in part because one of the prior art references discouraged such a combination). These factual issues surrounding Dow Jones' obviousness challenge cannot be resolved on this motion.
2. The Bobo Patent*fn2
The Bobo patent, U.S. Patent No. 5,675,507, has a priority date of April 28, 1995. It describes a system that allows users to access faxes by receiving HTML pages. SUF ¶ 51. The system allows the user to view faxes in text only, in text accompanied by a full-size image of the first page of the fax, or in text accompanied by a smaller (thumbnail) image of the first page of the fax. Id. ¶ 55. The user can also choose to see all of the pages of the fax in either full-size or thumbnail form. The program stores the user's preferences in a database. Id.
In my Markman order, I held that in the Claim 1 phrase "executing said set of functions to generate viewable content comprising said selected content data and formatting data," the term "formatting data" means HTML tags that specify the location of content, rather than its presentation (font, color, bold, etc.). See Dkt. 28, at 11. Ablaise contends that the Bobo patent does not meet this limitation because the patent "only describes a system for delivering content to the user without allowing the user to change the location of that content." Davis Decl., Ex. 10, ¶ 44.
Although the Bobo patent does tout its ability to accommodate a user's formatting preferences, see, e.g., id., Ex. 5, at 5:19-25, 5:59-60, 7:43-47, 10:38-42, there is a genuine factual dispute over whether it anticipates the "formatting data" element as I defined it in my Markman order. As Mr. Mulligan describes it, the Bobo patent allows the user to select the size of the fax image (thumbnail or full-size), and to select the number of fax images displayed (no image, image of the first page, or image of all pages). See Dkt. 95, Ex. 10, ¶ 62.
Ablaise argues persuasively that these are changes in the presentation of content or in the amount of content, rather than changes in the location of content. The user's selections may have an incidental impact on the location of text or images --just as increasing font size may push the selected text further down the page -- but my claim construction requires that the patent's HTML tags specify location, not simply affect it. Accordingly, summary judgment of anticipation is inappropriate as to the Bobo patent.
Even in Ablaise's view, there is only a narrow gap between the Bobo patent and Claim 1: while the Bobo patent uses HTML tags to affect the presentation of content, Claim 1 uses HTML tags to affect the location of content. Although obviousness from the Bobo patent is not Dow Jones' principal argument for invalidity, there are undisputed facts in the record that support such a finding.
Ablaise concedes that a person of ordinary skill in the art (even under its idea of the level of experience of such a person) would have been aware of HTML tags that affect content location. It does not dispute that "[b]y May 15, 1995, HTML, including the functions of the tags used therein, was well known to persons of skill in the art," SUF ¶ 5, nor does it dispute Dr. Bestavros' claim that "[m]any of the HTML tags in use prior to May 15, 1995 control[led] the location on the web page at which a certain piece of data (such as text or an image) [were] to be displayed," Dkt. 95, Ex. 7, ¶ 36. Indeed, Ablaise does not dispute Dow Jones' contention that Fishwrap used HTML tags to affect the location of content on the page.
The next question, then, is whether a person of ordinary skill in the art could have integrated location-changing HTML tags into the Bobo patent. Here, unlike with Fishwrap, there is undisputed evidence that such a modification to the Bobo patent would have been straightforward. For example, the Bobo patent employs the HTML "Image" tag, which is used to display an image on the page. See, e.g., Dkt. 95, Ex. 5, at 12:38. Ablaise concedes that an Image tag accompanied by an alignment tag (such as "align = middle") would meet the "formatting data" limitation of Claim 1. SUF ¶ 12. Therefore, a person of ordinary skill in the art would simply have to add an alignment tag to the existing image tag in the Bobo patent to meet all of Claim 1's limitations. "The combination of familiar elements according to known methods is likely to be obvious when it does no more than yield predictable results." KSR, 127 S.Ct. at 1739.
Lastly, there is the question of motivation. "When there is a design need or market pressure to solve a problem and there are a finite number of identified, predictable solutions, a person of ordinary skill has good reason to pursue the known options within his or her technical grasp." Id. at 1742. Ablaise's own expert notes that "[i]n the mid-1990's companies were rushing to establish an online presence and developers were working fast and furiously to get clients up on the Web." Davis Decl., Ex. 10, ¶ 65. Dr. Bestavros asserts, and Ablaise does not dispute, that "[b]etween 1990 and May 15, 1995, with the pervasiveness and wide acceptance of the web, there was a tremendous amount of technological development seeking to increase the usability of web pages and to bring the established features of various computer programs to the web platform."
Dkt. 85, Ex. 7, ¶ 37 (emphasis added). Both parties acknowledge that one of the "established features" of non-web programs was the capacity to personalize formatting. SUF ¶¶ 14-15. Taken together, these facts present clear and convincing evidence of both design need and market pressure to add location-changing HTML tags to the Bobo patent.
Ablaise has offered evidence of certain secondary indicia of non-obviousness: a long-felt but unmet need for the patented invention, skepticism about the possibility of the invention, and the commercial success of the invention. See Davis Decl., Ex. 10, ¶¶ 69-75. But Dow Jones argues, correctly, that Ablaise's evidence of long-felt need and skepticism in the industry is weak. See Dkt. 95, at 34-36. And while Ablaise has generated revenue from the '737 patent, it admits that it has done so through a coercive licensing scheme that has more to do with the costs of litigation than with the novelty of the patent. See Davis Decl., Ex. ¶ 90. In any event, "minimal indications of commercial success" are not cause to deny summary judgment if "obviousness [is] apparent from the prior art." Ball Aerosol and Specialty Container, Inc. v. Limited Brands, Inc., 555 F.3d 984, 994 (Fed. Cir. 2009). Such is the case here.
3. The Reuters (Risberg) Patent and the Sun Microsystems (Dasan) Patent
The Reuters patent (referred to as the "Risberg patent" by Ablaise) describes a "non-Web and non-Internet based software product that allows a user to customize the format and style of content on the display." SUF ¶ 84. The Sun Microsystems patent (referred to as the "Dasan patent" by Ablaise) describes a CGI program that allows a server to generate HTML-based pages dynamically based on a user's content preferences. SUF ¶ 87. Dow Jones argues that the combination of the two patents renders Claim 1 obvious.
The parties do not devote much energy to this theory, and neither will I. It suffices to say that there is a genuine factual dispute over whether either patent "contain[s] format identifiers, the selection of functions in dependence on received format identifiers, or the execution of functions in dependence upon a format identifier in order to change the location text/and or graphics on the page." Davis Decl, Ex. 10, ¶ 42. There is also a dispute over whether one of ordinary skill in the art could translate the Reuters patent's non-Web formatting approach to the web, given problems with "stateless connections, bandwidth limitations, latency in communication, lack of display control elements, and dependence upon the limitation of the HTML language." Id., ¶ 15. Accordingly, Dow Jones fails to meet its evidentiary burden for its theory.
C. The '530 patent
HTGrep is a program that performs keyword searches and returns results in an HTML-based web page. Dkt. 95, Ex. 7, ¶ 146. When a user request a search, the program allows the user to choose whether her search results will be displayed in paragraph form or in a numbered or bulleted list. SUF ¶ 61.
a. Prior Art
Dow Jones argues that HTGrep is prior art because it was "described in a printed publication" before May 15, 1995. 35 U.S.C. § 102(a). To establish that HTGrep meets this requirement, Dow Jones must show that the program was "made available to the extent that persons interested and ordinarily skilled in the subject matter or art exercising reasonable diligence [could] locate it and recognize and comprehend therefrom the essentials of the claimed invention." Bruckelmyer, 445 F.3d at 1378.
At his deposition, the inventor of HTGrep, Oscar Nierstrasz, claimed that he posted a link to the HTGrep source code on "numerous publically [sic] available newsgroups" on May 17, 1994. Davis Decl., Ex. 20, at 34:8-13. Though he could not recall the names of the newsgroups, he remembered that "various people downloaded [the source code] and contacted [him] with questions . . . [and] proposed changes." Id. at 47:7-8. To corroborate Nierstrasz's testimony, Dow Jones provides three postings from newsgroup subscribers. Two of the postings, written by subscribers to the "comp.infosystems.www" newsgroup, suggest HTGrep as an effective way to search HTML files. Id., Exs. 28, 29. The third, written by a subscriber to the "comp.lang.perl" newsgroup, indicates at least some familiarity with the HTGrep source code. Id., Ex. 30. All of the postings were made before May 15, 1995.
Ablaise presents no evidence to dispute Nierstrasz's testimony, and that testimony in combination with the three newsgroup postings is sufficient to carry Dow Jones' burden of proof. In In re Klopfenstein, 380 F.3d 1345 (Fed. Cir. 2004), the court held that a slide presentation that was printed, pasted onto poster boards, and displayed for two-and-a-half days at a academic conference was sufficiently publicly accessible to be considered a "printed publication" under 35 U.S.C. § 102. Id. at 1352. The court cited "the length of time the display was exhibited, the expertise of the target audience, the existence (or lack thereof) of reasonable expectations that the material displayed would not be copied, and the simplicity or ease with which the material displayed could have been copied," as the factors informing its conclusion. Id. at 1350. Each of those factors supports Dow Jones here: the source code was displayed for over a year before the priority date of the '530 patent; the two newsgroups (that we know of) to which Nierstrasz posted the link -- comp.infosystems.www and comp.lang.perl -- were for computer scientists and web programmers; and Nierstrasz encouraged newsgroup subscribers to copy and use his code.*fn3
Those factors compel a finding that HTGrep is prior art.
Ablaise argues first that the HTML tags that HTGrep uses are not "formatting type data." Citing the same portion of my Markman order that it cites in relation to the Bobo patent, Ablaise submits that HTGrep uses tags that only affect the presentation of content, rather than its location.*fn4 See Dkt. 28, at 11.
Ablaise's claim is unpersuasive. HTGrep uses tags such as "ol" and "ul," which determine whether an HTML list contains a bullet or a number to identify a particular item in the list. Dkt. 95, Ex. 10, ¶ 56. Inserting these tags not only places a bullet or a number before the selected text, but also indents the selected text. Davis Decl., Ex. 3, at 168-69. In this way, the tags are no different from the "center" and "p" tags that Ablaise admits are "formatting data" in the context of the '737 patent: each has the primary effect of shifting the selected text to a different location. When pressed on this point, Ablaise's expert could not offer a meaningful distinction between the HTGrep tags and those in Claim 1. See Davis Decl., Ex. 3, 180-85.
Ablaise's second argument is that HTGrep does not change between "a first and second specified page format as required by the claims." Dkt. 99, at 34. Ablaise does not expand on this argument, and its expert, Mr. Mulligan, provides a cursory and somewhat convoluted explanation of it in his deposition. See Dkt. 95, Ex. 3, at 187:4-19. HTGrep's capacity to generate content in simple paragraph form or in a bulleted or numbered list seems to clearly meet the two specified page format requirement of Claim 1. Accordingly, Dow Jones has presented clear and convincing evidence that HTGrep anticipates each element of Claim 1.
2. Thunderstone Bridge
Like HTGrep, Thunderstone Bridge was a program that generated search results in HTML form. It permitted the user to choose to view results in paragraph or non-paragraph format or in a vertical or horizontal orientation. SUF ¶ 79.
a. Prior Art
Dow Jones claims that Thunderstone Bridge is prior art because it was used before May 15, 1995. Ablaise admits that a version of Thunderstone was sold prior to May 15, 1995, but claims that there is not sufficient proof that the version sold had the relevant formatting functionality.
Ablaise is correct. John Turnbull, a Thunderstone corporate representative in 1994 and 1995, testified that he "believe[d]" that the version of Thunderstone sold before May 1995 offered the formatting features in question. See Dkt. 95, Ex. 19, at 40:3-9. That equivocal statement alone is not clear and convincing evidence of prior use. And Dow Jones cannot establish that the only document that discusses the Thunderstone program's formatting features -- the user manual -- described the product as it existed before May 15, 1995. The manual is dated October 2, 1996, id., Ex. 18, at 2, and though the chapter discussing the program's formatting features states it was last updated on May 2, 1995, id., at 327, there is a revision later in the chapter that is dated May 9, 1995, id., at 344. Without the user manual or some other documentary evidence to support Mr. Turnbull's testimony, Dow Jones fails to provide clear and convincing evidence that Thunderstone Bridge was prior art.
The Bobo patent renders Claims 1, 3, 4, and 6 of the '737 patent obvious, and the HTGrep program anticipates Claims 1-3 of the '530 patent. Accordingly, Dow Jones' motion for summary judgment [#95] will be granted by the accompanying order.