Netscape Communications Corp. v. Valueclick, Inc.

Full title: NETSCAPE COMMUNICATIONS CORP., Plaintiff, v. VALUECLICK, INC., et al.…

Court: United States District Court, E.D. Virginia

Date published: Apr 2, 2010

704 F. Supp. 2d 544 (E.D. Va. 2010)

Opinion

No. 1:09cv225.

April 2, 2010

Robert Leo Burns, II, Nicholas James Nugent, Umar Arshad, Finnegan Henderson Farabow Garrett Dunner LLP, Reston, VA, Charles Brandon Rash, Philip A. Riley, Finnegan Henderson Farabow Garrett Dunner LLP, Washington, DC, for Plaintiff.

Sheila Mary Costin, Amy Sanborn Owen, Richard Evans Craig, Cochran Owen LLC, Vienna, VA, Philip James Meitl, Bryan Cave LLP, Washington, DC, for Defendants.

MEMORANDUM OPINION

T.S. ELLIS III, District Judge.

In this patent infringement suit, plaintiff claims that defendants¹ wilfully infringed, and continue to infringe, U.S. Patent No. 5,774,670 ("the `670 patent"), colloquially known as the "Internet cookies" patent. Following full briefing and argument, the disputed patent claim terms were construed pursuant to Markman v. Westview Instruments, Inc., 517 U.S. 370 (1996). See Netscape Commc'ns Corp. v. ValueClick, Inc. (Netscape I), ___ F. Supp. 2d ___, 1:09cv225 (E.D. Va. Oct. 22, 2009) (Mem. Op.). The parties thereafter filed various cross-motions for summary judgment, which motions were partially resolved in a Memorandum Opinion dated January 29, 2010. See Netscape Commc'ns Corp. v. ValueClick, Inc. (Netscape II), ___ F. Supp. 2d ___, 1:09cv225 (E.D. Va. Jan. 29, 2010) (Mem. Op.) (hereinafter "the Memorandum Opinion").<sup>2</sup> At issue here is plaintiff's motion for reconsideration with respect to the entry of summary judgment in favor of defendants on a single issue, namely that claim 1 of the `670 patent is invalid under the statutory on-sale bar pursuant to 35 U.S.C. § 102(b).

1 The six named defendants are ValueClick, Inc., Mediaplex, Inc., FastClick, Inc., Commission Junction, Inc., MeziMedia, Inc., and Web Clients, L.L.C., (collectively "defendants").

2 Specifically, the parties filed various motions and cross-motions for summary judgment as to the following issues: (i) infringement and willful infringement; (ii) invalidity defenses pursuant to 35 U.S.C. §§ 102 and 112; (iii) equitable defenses of waiver, laches, and equitable estoppel; and (iv) preemption of defendants' state law counterclaims. These issues were resolved with the exception of plaintiff's motion concerning defendants' prior art invalidity defense, which was deferred pending further briefing.

I.

Under the Supreme Court's seminal decision in Pfaff v. Wells Electronics, Inc., 525 U.S. 55 (1998), a party invoking the § 102(b) statutory on-sale bar must prove (i) a commercial offer for sale of an invention that is (ii) ready for patenting prior to the statutory one-year period, i.e., the "critical date." The first Pfaff requirement is further separated into two distinct, constituent elements: (i) a commercial offer for sale, and (ii) a sufficient degree of identity between the subject of the offer for sale and the patented invention. See Allen Eng'g Corp. v. Bartell Indus., Inc., 299 F.3d 1336, 1353 (Fed. Cir. 2002). Importantly, both prongs of the Pfaff test must be proven by clear and convincing evidence. Clock Spring, L.P. v. Wrapmaster, Inc., 560 F.3d 1317, 1325 (Fed. Cir. 2009).

After considering the parties' pleadings, memoranda and exhibits submitted in support of their various motions for summary judgment, and the representations made by counsel in the course of the September 25, 2009 and December 18, 2009 hearings, the Memorandum Opinion found certain facts to be undisputed and material with respect to the on-sale bar analysis. See Netscape II, ___ F. Supp. 2d ___, 1:09cv225, at 5-9. Based on these factual findings, the Memorandum Opinion then found, by clear and convincing evidence, that claim 1 of the `670 patent was invalid under the § 102(b) on-sale bar because: (i) plaintiff had made a commercial offer to MCI Communications Corp. ("MCI") in September 1994; (ii) the subject of the offer was sufficiently identical to the method disclosed in claim 1; and (iii) the method disclosed in claim 1 was "ready for patenting" prior to October 6, 1994. Id. at 10-28.

In its motion for reconsideration, plaintiff argues that there was insufficient evidence of identity between the method disclosed in claim 1 and the subject of the offer to MCI, but does not challenge the fact of a commercial offer. Moreover, plaintiff contends that the invention was not "ready for patenting" prior to October 6, 1994. Notably, in moving for reconsideration plaintiff relies on some arguments rejected in the Memorandum Opinion, and additionally makes new arguments not raised in the earlier summary judgment briefs, including chiefly the application of the evidentiary corroboration requirement.³ The parties fully briefed and argued the issues at a March 26, 2010 hearing, at which time the motion was taken under advisement. Accordingly, the matter is ripe for disposition.

3 It is worth noting that defendants attach to their opposition brief a table identifying arguments that plaintiff argued in its original summary judgment briefs and now reargues in support of its motion for reconsideration. Defs.' Recons. Opp'n Br. Ex. A. Although defendants contend that all of plaintiff's current arguments were previously raised, this contention is inaccurate; it is clear that plaintiff raises new arguments and Federal Circuit authority not raised in its previous summary judgment briefs.

II.

At the threshold, the parties challenge the proper standard of review applicable to plaintiff's motion for reconsideration. Plaintiff moves for reconsideration under Rules 54(b), 59(e), and 60(b)(6), Fed.R.Civ.P. Motions for reconsideration pursuant to Rules 59(e) and 60(b)(6) are considered to be requests for an "extraordinary remedy" reserved only for "extraordinary circumstances" in which: (i) there is an intervening change in the law; (ii) new evidence not available at trial has been discovered; or (iii) a clear error of law must be corrected in order to prevent manifest injustice. See Pac. Ins. Co. v. Am. Nat'l Fire Ins. Co., 148 F.3d 396, 403 (4th Cir. 1998) (setting standards for motion pursuant to Rule 59(e)); Dowell v. State Farm Fire Cas. Auto. Ins. Co., 993 F.2d 46, 48 (4th Cir. 1993) (limiting relief under Rule 60(b)(6) to "extraordinary circumstances"). By contrast, a motion for reconsideration filed under Rule 54(b), Fed.R.Civ.P., ⁴ does not require a showing of extraordinary circumstances.

4 The Rule states, in pertinent part, that

any order or other decision . . . that adjudicates fewer than all the claims or the rights and liabilities of fewer than all the parties does not end the action as to any of the claims or parties and may be revised at any time before the entry of a judgment adjudicating all the claims and all the parties' rights and liabilities.

Rule 54(d), Fed.R.Civ.P.

The Fourth Circuit has made clear that where, as here, the entry of partial summary judgment fails to resolve all claims in a suit, Rule 54(d) — not Rule 59(e) or 60(b) — governs a motion for reconsideration:

[A]n order of partial summary judgment is interlocutory in nature. See, e.g., 11 Moore's Federal Practice § 56.40[3] (Matthew Bender 3d ed.) ("A partial summary judgment order is interlocutory. . . ."). Motions for reconsideration of interlocutory orders are not subject to the strict standards applicable to motions for reconsideration of a final judgment. See 12 Moore's Federal Practice § 60.23 ("Rule 60(b) does not govern relief from interlocutory orders. . . .").

Am. Canoe Ass'n v. Murphy Farms, Inc., 326 F.3d 505, 514-15 (4th Cir. 2003). And while doctrines such as law of the case guide a district court's discretion to reconsider an earlier partial summary judgment ruling, it is clear that "[t]he ultimate responsibility of the federal courts, at all levels, is to reach the correct judgment under law." Id.

Accordingly, the motion at bar is properly brought only pursuant to Rule 54(d), Fed.R.Civ.P., and therefore plaintiff is not required to make a showing of extraordinary circumstances. Instead, the goal here "is to reach the correct judgment under law." Id.

III.

Plaintiff first argues that the summary judgment record does not support a finding, based on clear and convincing undisputed evidence, that there is sufficient identity between the subject of the MCI offer for sale and claim 1 of the `670 patent. The Memorandum Opinion determined that the record evidence, considered as a whole, satisfied the clear and convincing standard because "the ultimate factfinder [would have] an abiding conviction that the truth of [the claimants'] factual contentions are highly probable." Pfizer, Inc. v. Apotex, Inc., 480 F.3d 1348, 1360 n. 5 (Fed. Cir. 2007) (citations and quotation marks omitted). Here, plaintiff contends: (i) that the Memorandum Opinion did not address whether two Markman claim term definitions — namely the "said http server" limitation and the sequence of steps disclosed in the claim 1 method — were offered to MCI; (ii) that the Memorandum Opinion incorrectly concluded that the "state object" claim limitation was offered for sale; (iii) that, to the extent testimonial evidence supports a finding of identity, evidence corroborating this testimony is lacking or insufficient; and (iv) that the record contains disputed material facts. These arguments are addressed in turn.

A.

In Netscape I, the claim term "said http server" was construed to mean "the same http server that received the request from the http client." Netscape I, ___ F. Supp. 2d ___, 1:09cv225, at 27-30. In addition, it was determined that the steps of the method disclosed in claim 1 should be construed, as follows: (i) request for a file by the http client; (ii) transfer of the file from the http server to the http client; (iii) transfer of the state object from the http server to the http client; and (iv) storage of the state object by the http client. See id. at 30-32. These Markman claim constructions were based chiefly on the patent's teaching that http servers are passive entities. See id. at 27, 31 (recognizing that passivity of http server is central to claim construction). Plaintiff correctly asserts that certain of these Markman claim term definitions were not explicitly addressed in the Memorandum Opinion. Nonetheless, clear and convincing undisputed record evidence supports the conclusion that these claim limitations, as defined in the Markman process, were included in the offer for sale. The dispositive principle here is inherency.

The doctrine of inherent anticipation is most commonly invoked with respect to prior art references, as in the context of a patent examiner's rejection of a patent application on the ground that the claimed invention was in the prior art, ⁵ or in the context of a claim of invalidity on grounds that a patented invention was used or described in a printed publication under 35 U.S.C. § 102(a)-(b).⁶ Yet, the Federal Circuit has extended the doctrine of inherent anticipation to the § 102(b) on-sale bar context, explaining that

5 See, e.g., In re Schreiber, 128 F.3d 1473 (Fed. Cir. 1999).

6 See, e.g., Perricone v. Medicis Pharm. Corp., 432 F.3d 1368 (Fed. Cir. 2005); Schering Corp. v. Geneva Pharms., Inc., 339 F.3d 1373 (Fed. Cir. 2003).

there is no requirement that the offer [for sale] specifically identify these limitations. Nor is there a requirement that [the inventor/assignee] must have recognized the significance of these limitations at the time of offer. If the process that was offered for sale inherently possessed each of the claim limitations, then the process was on sale, whether or not the seller recognized that his process possessed the claimed characteristics.

Scaltech Inc. v. Retec/Tetra, L.L.C., 178 F.3d 1378, 1384 (Fed. Cir. 1999) (internal citations omitted); see also Abbott Labs. v. Geneva Pharms., Inc., 182 F.3d 1315, 1319 (Fed. Cir. 1999). Although inherency is not proven simply by showing that something was a possible or even a probable consequence of a given set of circumstances, "if the natural result flowing from the operation of the process offered for sale would necessarily result in achievement of each of the claim limitations, then [the] claimed invention was offered for sale." Scaltech, 178 F.3d at 1384 (emphasis added). And notably, it is well-established that inherent anticipation is a factual question. See In re Schreiber, 128 F.3d 1473, 1477 (Fed. Cir. 1999); In re Buspirone Patent Antitrust Litig., 185 F. Supp. 2d 340, 360 (S.D.N.Y. 2002).⁷

7 Importantly, the Federal Circuit no longer requires proof that a person of ordinary skill in the art at the time would have recognized the inherent disclosure. See Schering Corp. v. Geneva Pharms., Inc., 339 F.3d 1373, 1377 (Fed. Cir. 2003) (explaining that Federal Circuit has vacated its decision in Elan Pharmaceuticals, Inc. v. Mayo Foundation for Medical Education Research, 304 F.3d 1221 (Fed. Cir. 2002), which required recognition of inherent disclosure).

As noted, the claim term "said http server," as defined in the Markman process, means "the same http server that received the request from the http client." Netscape I, ___ F. Supp. 2d ___, 1:09cv225, at 27-30. Although the Memorandum Opinion did not specifically address this defined claim term, there is no doubt that the cookies-functional Web browser technology sold to MCI in the summer and fall of 1994 required the same http server that received a file request to be the http server that returned state information in the form of a state object to the requesting http client. This characteristic is not simply an inference born of possibility or probability; rather, it is the "natural result" that necessarily follows from the fact that http servers are passive entities. See Scaltech, 178 F.3d at 1384. As stated in the Memorandum Opinion, the testimony of both Netscape and MCI executives confirms that Netscape sold a Web browser to MCI.⁸ Both plaintiff's and defendants' experts agree that the use of a Web browser by an http client to navigate the Internet necessarily and inherently requires an http client to send a request to an http server for a file, and for the http server to respond by transmitting a file which is then assembled and displayed on the client computer. See Pl.'s Ex. 4 ¶ 28; Defs.' Attach. 5 ¶ 25. Furthermore, it is undisputed that http servers are passive entities. As the `670 patent itself teaches, the http protocol governs the "architecture of the Web[, which] follows a conventional client-server model" in which "[the server] accepts commands from the client and cannot request the client to perform any action." `670 Patent Specification col. 1 ll. 60-62 (emphasis added). The parties' experts confirm this fundamental aspect of the http protocol, describing http client and http server interactions within the http protocol as a "request/response exchange," or as operating within a "request-response paradigm." See Pl.'s Ex. 4 ¶ 28; Defs.' Attach. 5 ¶ 25. From this, it necessarily follows that an http server will send a state object only in response to an http client request because an http server cannot, on its own and without a request, open a connection with an http client and send a state object. Put simply, under the http protocol it would be impossible for an http server to send a state object to an http client unless a request was first received.

8 See Defs.' Ex. 4, at 19, 96-99, 101 (deposition of Dr. John Klensin of MCI); Defs.' Ex. 5, at 61-62 (deposition of Dr. Vincent Cerf of MCI); Defs.' Ex. 19, at 101-02 (deposition of Marc Andreessen of Netscape).

Significantly, plaintiff, in arguing for reconsideration, points to no expert testimony or documentary evidence suggesting that an http server could, without being prompted by an http request, send a state object directly to an http client. And in fact, the record evidence points clearly and convincingly to a contrary conclusion, namely one that is consistent with the inherency of the "said http server" limitation. For instance, Netscape co-founder and vice president Marc Andreessen testified that the Netscape-MCI discussions focused sharply on "figur[ing] out a way to have the server be able to basically leave a piece of information on the client" in the context of "the server and browser . . . connect[ing] for the purpose of loading a page or session or whatever." Defs.' Opp'n Ex. 15, at 101-02 (emphasis added). This testimony establishes Andreessen's understanding: (i) that the to-be-written MCI software would operate in the context of a single http server accepting a request to connect with, and transmit a file to, an http client; and (ii) that MCI required computer source code directing a single http server "to basically leave a piece of information on the client" — i.e., the state object containing state information — while the client and server remained connected. Significantly, Andreessen describes this as being the essence of MCI's design requirements; his testimony does not disclose the presence or participation of a second http server that does not receive a request from the http client. Accordingly, the inherency of the "said http server" limitation in the Web browser sold to MCI is necessarily dictated by the nature of the http environment and confirmed by Andreessen's undisputed testimony.

The same conclusion obtains with respect to the sequence of steps disclosed in claim 1. Again, no evidence in the record disputes the fact that an http server is a passive entity that responds to an http client's request. Accordingly, the process must begin with a request by an http client to an http server. This result is the natural and necessary consequence of the operation of the Web browser sold to MCI. See Pl.'s Ex. 4 ¶ 28 (describing "response/request exchange"); Defs.' Attach. 5 ¶ 25 (describing "response-request paradigm"). It follows that the requested file and a state object be sent next, prior to the storage of the state object on the client computer. As Dr. John Klensin of MCI testified in his deposition, "the state information where the user was last in the marketplace needed to be transmitted from the server to the client in some way. The client needed to keep it." Defs.' Ex. 4, at 98-99 (emphasis added). As noted in the Memorandum Opinion, this testimony tracks precisely steps 3 and 4 of claim 1 — that is, the transmission of state information in the form of a state object and the storage of the state object — and plaintiff points to no evidence suggesting that a state object might be stored before it is received by the http client. Indeed, such an argument would be nonsensical. Although Dr. Klensin states that during the August 10, 1994 meeting the participants did not discuss "in which sequence of the HTTP operations these things would be done, which sequence of the client and server,"⁹ the fact that this limitation was not specifically identified during the August 10, 1994 is immaterial. As the Federal Circuit has explained, "[w]e note that there is no requirement that the offer specifically identify these limitations" provided that "the process that was offered for sale inherently possessed each of the claim limitations." Scaltech, 178 F.3d at 1384. In this case, the nature of the http protocol — specifically the passive character of servers — necessarily and inherently requires that the Web browser sold to MCI perform the steps of the claimed process in the order as construed in Netscape I.

9 Defs.' Ex. 4, at 96.

In sum, the undisputed summary judgment record establishes that, in September 1994, Netscape sold a cookies-functional Web browser to MCI before the critical date. Also clear from the undisputed record is the fact that this Web browser operated within the limitations imposed by the http protocol, including the fact that http servers are passive entities that interact with an http client only on an http client's request. Therefore, it necessarily follows — and is not simply a matter of possibility or probability — that the Web browser sold to MCI prior to the critical date inherently satisfies the "said http server" and the sequence of steps limitations disclosed in claim 1 of the `670 patent.

B.

Next, plaintiff argues that no evidence proves, as the Memorandum Opinion concluded, that the state object was part of the offer for sale to MCI in August and September 1994. This argument is wholly unpersuasive. In addition to the fact that state information is inherently transferred in a structured manner such that it forms a state object — that is, "data having a predetermined structure that specifies state information"¹⁰ — ample record evidence demonstrates that the state object was in fact made part of the Web browser sold to MCI prior to the critical date.¹¹ This follows inexorably from the fact that the term "cookie" in the http context is synonymous with the term "state object."¹² Andreessen testified at least three times during his deposition that the cookies invention was used to satisfy MCI requirements.¹³ This testimony is corroborated by John Giannandrea, a Netscape senior programmer who later served as the Web browser division's chief technology officer, and Lou Montulli, the Netscape programmer who invented the cookies technology and filed the `670 patent application. Specifically, Montulli confirms that he invented the cookie mechanism in response to MCI's design requirements in the summer of 1994. See Netscape II, ___ F. Supp. 2d ___, 1:09cv225, at 20-21 (finding clear and convincing undisputed evidence relating to shopping cart and commerce server project in summer 1994). Giannandrea's deposition testimony makes clear that the invention and implementation of the cookies technology occurred prior to the critical date, namely during a series of July and August 1994 meetings in which he and Montulli discussed "the idea that the Web server would request that the browser send the cookie back to the Web server." Defs.' Ex. 21, at 60.

10 As explained in the Memorandum Opinion, "claim I neither requires nor specifies that a particular source code must be used," and "requires only that state information be conveyed in a structured manner." Because "the storage of state information on a client necessarily and inherently requires that this data be conveyed in a structured manner," it follows that the August 10, 1994 discussion of state information storage on an http client inherently anticipates claim 1 under Scaltech. Netscape II, ___ F. Supp. 2d ___, 1:09cv225, at 18-19.

11 As noted supra, the fact that the state object was not identified during the August 10, 1994 meeting as the solution to statelessness is irrelevant because "there is no requirement that the offer specifically identify these limitations." Scaltech, 178 F.3d at 1384. More fundamentally, this argument fails because it is undisputed, as discussed in infra, that Netscape ultimately decided to use a state object in creating Web browser technology for MCI.

12 See Defs.' Opp'n Ex. 16, at 120 (Montulli agreeing that the "state object is called a cookie"); Defs.' Attach. 4 ¶ 16 (defendant's expert concluding that "[t]he cookie is a short hand term for state object. . . ."); see also `670 Patent Specification col. 7 ll. 13-18 (teaching that, in one embodiment, a piece of state information "is referred to as a `cookie'").

13 See, e.g., Pl.'s Ex. 45, at 102 (explaining that Netscape's developing software for marketplaceMCI "eventually led to the creation of cookies"); Defs.' Opp'n Ex. 15, at 102 (recalling that "cookie" referred to "information that the server could leave on the client" and that this process would satisfy MCI's design requirement); id. at 140 (agreeing that "the cookie functionality would be included in browsers sold not just to MCI, but to other customers as well").

Added support for this conclusion is found in the Netscape-MCI software licensing agreement, which provides the terms by which Netscape would license "client and server software . . . in connection with such electronic shopping mall and to provide access to and browsing capabilities with respect to the World Wide Web and the Internet generally." Defs.' Ex. 13, at 1. Contrary to plaintiff's assertion, the document expressly discloses that the Web browser sold to MCI was capable of performing the claimed method, as it explicitly discusses the storage of "cookies" on a client computer.¹⁴ Although the agreement was executed in February 1995, it nonetheless provides documentary evidence corroborating other testimonial evidence on the state object's sale prior to the critical date. There is no dispute that the agreement relates to software technology that Netscape and MCI began discussing on August 10, 1994. Indeed, plaintiff concedes as much when it cites the agreement, albeit inaccurately, in both its summary judgment and motion for reconsideration briefs as evidence that "MCI's high-level system requirements with respect to state information was limited to . . . state information stor[age] on the client side,"¹⁵ the central design requirement discussed on August 10, 1994.¹⁶ Moreover, Article 6(b)(i) of the agreement indicates: (i) that MCI made a $500,000 prepayment for "Netscape Software" on September 23, 1994; and (ii) that MCI would pay $2,400,000 either by January 1, 1995, or on receipt of "Version 1.A" of the client and server software. Defs.' Ex. 13, at 9. This evidence plainly shows that the agreement records, and relates to, events that occurred prior to its February 1995 execution. Put another way, the agreement memorializes the Netscape-MCI relationship — which began in August 1994 and led to a $500,000 prepayment in September 1994 — all concerning the development, implementation, and delivery of software technology that would, as the agreement states, fulfill MCI's "desire[] to develop an electronic shopping mall."¹⁷ Defs.' Ex. 13, at 1. Thus, it is clear from the summary judgment record that a Web browser equipped to receive and store a state object, i.e., a "cookie," was sold or offered for sale to MCI prior to the critical date.

14 See Defs.' Ex. 13 Attach. A at 2 ("Support for multiple (10) Client Cookies to handle shopping carts/discrete one-time one-line info deliver, etc."); id. Attach. A-1 at 1.18 ("The number of cookies that can be stored in a client needs to be increased. . . ."); see also id. Attach. A at 3 ("Automatic transparent identification of client.").

15 Pl.'s Summ. J. Statement of Disputed Facts ¶ 3; see also Pl.'s Recons. Br. at 14 n. 45. Specifically, the statement is inaccurate because the agreement explicitly refers to the storage of "cookies" — i.e., a state object — on a client computer, and is not, contrary to plaintiff's contention, limited to the "high-level" requirement of state information storage on a client computer. Indeed, plaintiff's brief in support of summary judgment recognizes that the February 1995 agreement "discuss[es] `cookies.'" Pl.'s Summ. J. Br. at 18.

16 See Netscape II, ___ F. Supp. 2d ___, 1:09cv225, at 16-17.

17 As explained in the Memorandum Opinion, clear and convincing evidence establishes that MCI negotiated with Netscape in the summer of 1994 to develop software for implementation in marketplaceMCI, an online shopping mall. See Netscape II, ___ F. Supp. 2d ___, 1:09cv225, at 5, 20-21 (discussing elements of marketplaceMCI and Montulli's inventing the cookies technology for a "shopping cart methodology" to be used in an "e-commerce server"). It is worth noting that Michael Mael, an MCI executive involved in the Netscape-MCI discussions, testified in his deposition that a dispute between Netscape and the University of Illinois over the use of certain confidential Web browser technology raised substantial concern within MCI in the fall of 1994. In response, MCI advised Netscape that further payments would not be made until the dispute was resolved. See Defs.' Ex. 11, at 83. Yet, whether this testimony explains the delay between the September 1994 offer for sale and the execution of the February 1995 agreement is not material.

In support of its motion for reconsideration, plaintiff argues that the Web browser sold to MCI did not necessarily, and thus did not inherently, use the claimed cookies invention. To that end, plaintiff cites Montulli's testimony that "there were other ways of dealing with that problem, but none of them were terribly elegant." Defs.' Opp'n Ex. 16, at 119. Yet, this statement, when placed in context, does not undermine the conclusion that the state object was inherent to the transfer and storage of state information on a client computer. Although Montulli at one point asserted that there were other ways of addressing statelessness, the sole alternate solution identified by Montulli in his deposition — namely the http server's sending a unique, recognizable Web address to the http client — does not require the http client to store state information, and thus would not have satisfied MCI's specific design requirements. See id. In other words, Montulli's statement casts no doubt on the inherency of a state object because his testimony does not disclose that there were other ways of storing state information on a client computer. Instead, he stated that there were other ways of addressing statelessness generally, and identified one such example that did not require storage of state information on a client computer. Thus, the evidence cited by plaintiff does not demonstrate that the use of a state object was only one of a number of means by which Montulli could have satisfied MCI's specific requirement of storing state information on a client computer, a requirement distinct from addressing http statelessness more generally.¹⁸ Accordingly, this evidence does not call into question the conclusion that the state object was part of the offered technology.

18 Even assuming that the state object was not inherent to the storage of state information on a client computer, plaintiff's argument still fails because it ignores evidence that Netscape and Montulli actually used the cookies technology to satisfy MCI requirements. As the Federal Circuit has explained in the context of an anticipating prior art reference, the doctrine of inherent anticipation may be asserted in circumstances where "the reference is silent about the asserted inherent characteristic," making it necessary to fill "such a gap . . . with recourse to extrinsic evidence." See Continental Can Co. USA, Inc. v. Monsanto Co., 948 F.2d 1264, 1268 (Fed. Cir. 1991) (emphasis added). In this case, the conclusion that claim 1 was offered for sale prior to the critical date does not rest on inherency alone; rather, direct evidence — namely the deposition testimony of Andreessen, Giannandrea, and Montulli, and the Netscape-MCI software licensing agreement — intrinsically establishes that the state object, i.e., a "cookie," was offered for sale prior to the critical date.

C.

The parties spill much ink on whether, and to what degree, evidence supporting a finding of invalidity must be corroborated by either testimonial or documentary evidence. Thus, it is appropriate to clarify the corroboration requirement.

Confusion concerning the corroboration requirement stems from two arguably contradictory decisions. First, in Thomson, S.A. v. Quixote Corp., 166 F.3d 1172 (Fed. Cir. 1999), the Federal Circuit affirmed a jury verdict, based on the testimony of two putative non-party inventors, that the patent was invalid under 35 U.S.C. § 102(g). Distinguishing its decision in Woodland Trust v. Flowertree Nursery, Inc., 148 F.3d 1368 (Fed. Cir. 1998), the Federal Circuit in Thomson held that the corroboration requirement only applied "where the inventor is self-interested in the outcome of the trial and is thereby tempted to `remember' facts favorable to his or her case." Thomson, 166 F.3d at 1175-76. From this, it followed that corroboration was only required when the testifying inventor was asserting a claim of derivation or priority of his invention and was (i) a named party, (ii) an employee of, or assignor to, a named party, or (iii) otherwise in a position to gain directly and substantially from the invalidation of the contested patent claims. Id. at 1176.

Thomson was distinguished — but not expressly overruled — by a subsequent panel decision in Finnigan Corp. v. Int'l Trade Comm'n, 180 F.3d 1353 (Fed. Cir. 1999). In Finnigan, the Federal Circuit stated that testimonial evidence alone rarely satisfies the clear and convincing standard because "it is rare indeed that some physical record . . . does not exist." Id. at 1366-67; see also Sandt Tech., Ltd. v. Resco Metal Plastics Corp., 264 F.3d 1344, 1350-51 (Fed. Cir. 2001) (expressing, in dictum, that testimonial evidence, as contrasted with documentary evidence, "provides the most reliable proof" of corroboration). And while the corroboration requirement was historically applied to prior public use disputes under § 102(a) and (b), inventorship disputes under § 102(f), and priority disputes under § 102(g), the Federal Circuit saw no reason to apply a different rule to other subsections of § 102. See Finnigan, 180 F.3d at 1367 n. 10.¹⁹ Taking aim at Thomson, which essentially applied the corroboration requirement based on the purported inventor's level of interest and relationship to the case, the Federal Circuit in Finnigan expressly held that "corroboration is required of any witness whose testimony alone is asserted to invalidate a patent, regardless of his or her level of interest." Id. at 1369. This result was consistent with Thomson, the Federal Circuit explained, principally because Thomson (i) addressed the sufficiency, not the necessity, of corroboration, and (ii) involved both documentary and testimonial evidence.

19 Citing Finnigan, the Federal Circuit in 2003 expressly held that "[c]orroboration is also required where, as here, the testimony is from an accused infringer concerning the sale (or offer to sell) or public use of an invention more than one year before the filing of a patent." Lacks Indus. v. McKechnie Vehicle Components USA, Inc., 322 F.3d 1335, 1349 (Fed. Cir. 2003).

One district court has noted in discussing Thomson and Finnigan, that "[i]t is not at all that easy to understand why patent invalidity, of all things, should rank with treason in terms of the need for special corroboration." Engate, Inc. v. Esquire Deposition Servs., L.L.C., 331 F. Supp. 2d 673, 685 (N.D. Ill. 2004). Consequently, courts — including the Federal Circuit — have applied the corroboration requirement somewhat inconsistently. On the one hand, the Federal Circuit has affirmed a decision invalidating patents based solely on the testimony of one witness. See Eisenberg v. Alimed, Inc., 2000 U.S. App. LEXIS 19121 (Fed. Cir. Aug. 8, 2000). On the other hand, the Federal Circuit has emphasized that "a patent cannot be invalidated based on one person's testimony alone without corroborating evidence, particularly documentary evidence." Adenta GmbH v. OrthoArm, Inc., 501 F.3d 1364, 1371 (Fed. Cir. 2007). Yet, it seems that courts have increasingly concluded that Thomson has been overruled by implication, and therefore the Finnigan corroboration requirement applies in all circumstances. As one district court has observed,

We agree with Judge Sachs of the Western District of Missouri who, in deciding a summary judgment motion, said that even though Thomson has not been overruled, in light of all the "subsequent decisions [that] have either returned to the rule as it was enunciated before Thomson, see Oney v. Ratliff, 182 F.3d 893, 896-97 (Fed. Cir. 1999), or have distinguished Thomson under circumstances suggesting that its holding may not be viable, see Finnigan, 180 F.3d at 1367-68," the Court's "safest" option is "to operate under the assumption that the corroboration requirement still exists." K K Jump Start/Chargers, Inc. v. Schumacher Elec. Corp., 82 F. Supp. 2d 1012, 1018 n. 7 (W.D. Mo. 2000). One can only hope that the Federal Circuit will resolve the conflict created by its decisions on this issue.

See Engate, 331 F. Supp. 2d at 685. This observation is persuasive. Until the Federal Circuit further clarifies this issue, the safest course, in the circumstances, is to apply the Finnigan corroboration requirement broadly, including to on-sale bar claims.

Despite the confusion concerning when the corroboration requirement applies, the manner of its application is well-settled. Put another way, the principles that guide a court's evaluation of the sufficiency of corroborating evidence are clear. The seminal decision in this regard is Woodland Trust v. Flowertree Nursery, Inc., 148 F.3d 1368 (Fed. Cir. 1998). In that case, the Federal Circuit endorsed the following factors first identified by its predecessor institution, the U.S. Court of Customs and Patent Appeals, in In re Reuter, 670 F.2d 1015, 1021 n. 9 (C.C.P.A. 1981):

(1) the relationship between the corroborating witness and the alleged prior user;

(2) the time period between the event and trial;

(3) the interest of the corroborating witness in the subject matter of the suit;

(4) contradiction or impeachment of the witness' testimony;

(5) the extent and details of the corroborating testimony;

(6) the witness' familiarity with the subject matter of the patented invention and the prior use;

(7) probability that a prior use could occur considering the state of the art at the time; and

(8) impact of the invention on the industry, and the commercial value of its practice.

Woodland Trust, 148 F.3d at 1371 (quoting Price v. Symsek, 988 F.2d 1187, 1195 n. 3 (Fed. Cir. 1993)).

These factors are notable in two respects. First, although the witness's interest in the litigation is irrelevant to whether the corroboration requirement applies in the first instance under Finnigan, the witness's interest is relevant to the sufficiency of corroboration. Second, the Woodland Trust factors expressly contemplate that an inventor's testimony may be corroborated by other testimony, and thus invalidity may be determined even in the absence of documentary evidence. See TypeRight Keyboard Corp. v. Microsoft Corp., 374 F.3d 1151, 1159 (Fed. Cir. 2004) ("Both physical evidence and oral testimony of a disinterested party can serve to satisfy the corroboration requirement."); see also Engate, 331 F. Supp. 2d at 685-86 (rejecting argument that "testimonial evidence from multiple witnesses alone is never sufficient to invalidate a patent"). At bottom, the Woodland Trust factors reflect a "rule of reason" analysis, by which a "court examines all pertinent evidence" — including circumstantial evidence — "to determine the credibility of the inventor's story." Brown v. Barbacid, 276 F.3d 1327, 1335 (Fed. Cir. 2002) (citations and internal quotation marks omitted). And contrary to plaintiff's unsupported argument that "Federal Circuit authority insists that there still be corroboration for each and every limitation,"²⁰ the Federal Circuit's predecessor institution, the U.S. Court of Customs and Patent Appeals, explained in the context of an interference proceeding that "[t]he law does not impose an impossible standard of `independence' on corroborative evidence by requiring that every point of a reduction to practice be corroborated by evidence having a source totally independent of the inventor" because "such a standard is the antithesis of the rule of reason." Knorr v. Pearson, 671 F.2d 1368, 1374 (C.C.P.A. 1982).²¹ Similarly, the Federal Circuit more recently expressed the same view in rejecting a patentee's "attempts to discredit each piece of evidence, one by one, to show that clear and convincing evidence invalidating the patent was lacking" on a motion for judgment as a matter of law. Adenta, 501 F.3d at 1372. Accordingly, the Federal Circuit's decisions make clear that the corroboration requirement's fundamental aim is to determine whether the "testimony of the witnesses together with the documentary evidence provide[] a coherent and convincing story." Id. at 1371.

20 Pl.'s Recons. Br. at 9.

21 The decisions of the Court of Customs and Patent Appeals govern here pursuant to South Corp. v. United States:

As a foundation for decision in this and subsequent cases in this court, we deem it fitting, necessary, and proper to adopt an established body of law as precedent. That body of law represented by the holdings of the Court of Claims and the Court of Customs and Patent Appeals announced before the close of business on September 30, 1982 is most applicable to the areas of law within the substantive jurisdiction of this new court. It is also most familiar to members of the bar. Accordingly, that body of law is herewith adopted by this court sitting in banc.

S. Corp. v. United States, 690 F.2d 1368, 1370 (Fed. Cir. 1982) (in banc); see also 8-21 Chisum on Patents § 21.02 (discussing Federal Circuit's adoption of predecessor courts' holdings).

These principles, applied here, compel the conclusion that the summary judgment record satisfies the Finnigan corroboration requirement under the factors enumerated in Woodland Trust. The evidence in the summary judgment record pertaining to the identity issue is found in: (i) deposition testimony of Andreessen, Giannandrea, and Montulli of Netscape; (ii) deposition testimony of Drs. Vincent Cerf and Klensin of MCI; (iii) reports by plaintiff's and defendants' experts; (iv) the Netscape-MCI software licensing agreement; and (v) the draft client-side cookies source code entered into the Netscape source code repository on October 3 or 4, 1994.²² To begin with, the record evidence establishing identity between the matter sold or offered for sale and claim 1 is largely from plaintiff's executives and programmers, persons who presumably have an interest in defeating defendants' identify claim. Notwithstanding this, their testimony confirms, rather than undermines, the conclusion that claim 1 was offered for sale prior to the critical date. Most probative are Andreessen's and Montulli's statements, taken together, that clearly and convincingly establish that the invention of cookies technology was a direct result of MCI's demand for storage of state information on a client computer.²³

22 Plaintiff argues that the draft cookies source code cannot serve as corroborating evidence because defendants have not submitted any expert testimony deciphering the source code. Yet, the absence of a defendants' expert on this point is immaterial because two of plaintiff's experts have analyzed the source code. Notably, the conclusion reached by plaintiff's experts is clear and undisputed: computer code pertaining to Web browser cookie functionality existed as of October 3 or 4, 1994. See Pl.'s Ex. 42 ¶¶ 22-31 (expert report of Edwin Aoki); Pl.'s Ex. 43 ¶¶ 35-47 (expert report of Judson Valeski).
In this regard, it is also worth noting defendants' argument that the summary judgment record contains other corroborating evidence, namely the alpha version of the software sent to MCI in September 1994. Yet, this evidence does not affect or alter the corroboration analysis because, as noted in the Memorandum Opinion with respect to defendants' public use invalidity defense, "the record evidence does not indicate whether the `alpha' or `beta' versions of plaintiff's Web browser sent to [MCI] were cookies-functional." See Netscape II, ___ F. Supp. 2d ___, 1:09cv225, at 30. Accordingly, this evidence neither confirms nor refutes the other record evidence relating to identity.

23 See Pl.'s Ex. 45, at 102 (Andreessen testifying that the development of software for MCI "eventually led to the creation of cookies"); Defs.' Opp'n Ex. 15, at 140 (Andreessen testifying that the "cookie functionality would be included in browsers sold . . . to MCI"); Defs.' Ex. 16, at 114-19 (Montulli recounting that he invented the Web cookie in response to a client demand for a "shopping cart methodology" that would address the "sessionless nature of HTTP" within the context of an "e-commerce server").

Moreover, although plaintiff asserted in the course of oral argument that Dr. Klensin is an "interested party" because he is a retained expert, this fact does not make him an "interested party" as understood by the Federal Circuit. More precisely, Dr. Klensin is neither a named party, nor an employee of, or assignor to, a named party, nor a person who is otherwise in a position to gain directly and substantially from the invalidation of the contested patent claims. See Thomson, 166 F.3d at 1175-76. Yet, even assuming arguendo that Dr. Klensin is an interested party, any concern that Klensin's testimony may be biased in favor of invalidity is mitigated by the fact that his testimony with respect to claim 1 is consistent with the testimony of plaintiff's witnesses. See Adenta, 501 F.3d at 1372 ("While the court determined that Heiser and Schendell were clearly interested parties, as they were named parties to the suit, their testimony was corroborated by the testimony of [non-interested parties] Voudouris, Russell, and Tuneberg."). Indeed, Dr. Klensin, Dr. Cerf, and Andreessen recount the same specific sequence of events that ultimately culminated in the sale of the claim 1 method to MCI: (i) Netscape first proposed that state information be stored on the http server; (ii) MCI rejected this proposal and required state information to be stored on the client computer; and (iii) Netscape and MCI clearly understood by the August 10, 1994 meeting's conclusion that state information would be stored on a client computer.²⁴ These undisputed facts are central to the conclusion that claim 1 was offered for sale prior to the critical date.

24 See Defs.' Ex. 4, at 19, 96-99, 101 (deposition of Dr. John Klensin of MCI); Defs.' Ex. 5, at 61-62 (deposition of Dr. Vincent Cerf of MCI); Defs.' Ex. 19, at 101-02 (deposition of Marc Andreessen of Netscape).

This testimonial evidence is further corroborated by undisputed documentary evidence. Specifically, the Netscape-MCI software licensing agreement makes clear that MCI purchased Web browser technology capable of receiving and storing a cookie — a synonym for a state object — and that MCI paid Netscape a $500,000 prepayment for this technology on September 23, 1994.²⁵ See Defs.' Ex. 13, at 9. The draft cookies source code entered into the Netscape source code repository on October 3 or 4, 1994, as interpreted by plaintiff's experts, also confirms that the cookies technology was being implemented prior to the critical date. See Pl.'s Ex. 42 ¶¶ 22-31 (expert report of Edwin Aoki); Pl.'s Ex. 43 ¶¶ 35-47 (expert report of Judson Valeski).

25 It is worth emphasizing, as discussed supra, that although the agreement was executed in February 1995, it is clear that the agreement records and memorializes events that occurred prior to the critical date concerning the sale of cookies technology to MCI for implementation in marketplaceMCI.

In response, plaintiff cites the Federal Circuit's decision in Adenta GmbH v. OrthoArm, Inc. In Adenta, the Federal Circuit held that OrthoArm's motion for judgment as a matter of law as to invalidity was correctly denied on the ground that the trial record contained ample corroborating evidence to support the jury's invalidity verdict. See 501 F.3d at 1370-73. Specifically, the Federal Circuit held that the following evidence was sufficient to allow the jury to find public use by clear and convincing evidence: (i) testimony by five witnesses, both interested and disinterested, that the invention was displayed at a trade show; (ii) a transmitted fax containing drawings of the invention; (iii) a German patent application disclosing an embodiment of the invention; (iv) a transmitted fax requesting the display of the invention; (v) an article containing photographs of the invention; and (vi) a letter from Adenta's lawyers recognizing that the invention had been displayed at the trade show.²⁶ See id. Plaintiff argues that invalidity can only be proven by clear and convincing evidence where the same quantum of evidence presented in Adenta exists. Yet, Adenta stands for no such proposition. Although the evidence presented in that case was sufficient to meet the clear and convincing standard, the Federal Circuit did not require, and no inference can be drawn, that the same quantum of evidence is necessary in every case. Put simply, Adenta does not instruct courts to conduct a checklist-like analysis. Instead, Adenta itself recognizes that the rule of reason analysis under Woodland Trust contemplates a case-by-case weighing of the evidence in the particular circumstances presented, using a multi-factor approach. See id. at 1372 ("OrthoArm attempts to discredit each piece of evidence, one by one, to show that clear and convincing evidence invalidating the patent was lacking. However, the jury considered all the evidence."); accord Sandt Tech., 264 F.3d at 1350 ("Each corroboration case must be decided on its own facts with a view to deciding whether the evidence as a whole is persuasive." (citation and quotation marks omitted)). In the final analysis, it is only necessary that the "testimony of the witnesses together with the documentary evidence provide[] a coherent and convincing story." Adenta, 501 F.3d at 1371. In this case there can be no doubt that the consistent testimony of Netscape and MCI executives and programmers, which is further corroborated by documentary evidence, tells a "coherent and convincing story" concerning the sale of claim 1 to MCI prior to the critical date, thereby satisfying the corroboration requirement under the Woodland Trust factors.

26 Notably, Adenta addressed the application of the corroboration requirement on a motion for judgment as a matter of law following trial on the invalidity defense and the jury's resolution of a factual dispute. Accordingly, the question in Adenta was whether there was sufficient evidence on which a jury could find invalidity clearly and convincingly. See Adenta, 501 F.3d at 1370-73. By contrast, in this case, as discussed infra, the relevant evidence concerning the sale of claim 1 prior to the critical date is consistent and uncontradicted, and thus the issue need not be submitted to a jury because no material factual dispute exists.

D.

Finally with respect to identity, Netscape contends that the summary judgment record contains only disputed evidence concerning whether the offer of sale to MCI prior to the critical date met the limitations found in claim 1 of the `670 patent. Distilled to its essence, Netscape's argument is that while the storage of state information on an http client was discussed at the August 10, 1994 meeting, the remaining claim 1 limitations were not identified, and thus claim 1 was not offered for sale prior to October 6, 1994. This argument fails on legal and factual grounds.

To begin with, it is well-established, as discussed supra, "that there is no requirement that the offer specifically identify these limitations," provided that "the process that was offered for sale inherently possessed each of the claim limitations." Scaltech, 178 F.3d at 1384. Furthermore, as also discussed supra, clear and convincing record evidence discloses that the limitations purportedly missing from the offer for sale were in fact inherent to — and in fact made a part of — the Web browser technology sold to Netscape prior to the critical date.

Plaintiff's cited evidence neither compels a contrary conclusion nor creates a dispute of material fact. Specifically, Dr. Cerf's statement that "sometime after August, several months after August, [Netscape] produced a product that [MCI was] able to validate performed the way that [MCI] needed it to"²⁷ is irrelevant to the on-sale bar analysis. As the Federal Circuit has explained, even assuming arguendo that Netscape delivered the product to MCI after October 6, 1994, "[i]t is immaterial that the record shows no delivery . . . until after the critical date." Weatherchem Corp. v. J.L. Clark, Inc., 163 F.3d 1326, 1333 (Fed. Cir. 1998).

27 Pl.'s Opp'n Br. Ex. 2, at 61-63.

Moreover, Dr. Cerf's additional statement that "there is some distance between [the August 10, 1994 discussion of storing state information on the client] and the specifics and the details that are found in the patent"²⁸ does not create a material dispute of fact concerning the specificity of the Netscape-MCI discussions that precludes entry of summary judgment in favor of defendants as to claim 1. As explained in the Memorandum Opinion, the parties' disagreement as to whether specific attributes of the general method disclosed in claim 1, such as name-value pairs or secure encryption, were discussed at the August 10, 1994 meeting is relevant only to claims 2-8 and 15-26; the parties' disagreement is irrelevant to the general four-step method disclosed in claim 1. See Netscape II, ___ F. Supp. 2d ___, 1:09cv225, at 22 (citing disputed evidence and finding that summary judgment cannot be entered in favor of defendants on the other contested patent claims). Nor does Dr. Cerf's statement preclude a finding that claim 1 was offered for sale prior to the critical date. Indeed, record evidence of events occurring between August 10 and October 6, 1994, proves otherwise. More precisely, the record evidence, taken as a whole, establishes clearly and convincingly that discussions and negotiations between Netscape and MCI began on August 10, 1994, and continued through the fall of 1994. See id. at 10-14 (detailing negotiations and finding offer for sale as of early September 1994); see also Defs.' Ex. 5, at 82-83 (Dr. Cerf confirming follow-up meetings with Netscape); Defs.' Ex. 8, at 64 (Netscape vice president James Sha testifying that he had many subsequent meetings with MCI personnel after August 10, 1994). Evidence of what was offered for sale is not restricted to the August 10, 1994 meeting, as Netscape implicitly argues. Instead, as discussed supra and in the Memorandum Opinion, the undisputed record evidence — which includes deposition testimony by both Netscape and MCI employees, who participated in the August 10, 1994 meeting and/or thereafter developed the software for MCI — establishes clearly and convincingly: (i) that MCI sought software technology that facilitated storage of state information on a client computer; (ii) that Netscape sold Web browser technology to MCI capable of storing state information on a client computer by September 1994; and (iii) that Montulli invented the cookies technology in the summer of 1994 specifically to satisfy MCI's design requirements, and wrote computer source code implementing the invention by October 3 or 4, 1994.²⁹

28 Pl.'s Ex. 2, at 136-37.

29 It is worth briefly addressing plaintiff's two additional arguments. First, the fact that Dr. Klensin's calendar shows that "[Netscape] decides they can support name-value pairs" on October 10, 1994, is irrelevant to whether claim 1 was offered for sale. Name-value pairs is a limitation found in claim 2, not in claim 1. Second, plaintiff's citation to Plumtree Software, Inc. v. Datamize, LLC, 473 F.3d 1152 (Fed. Cir. 2006), is wholly inapposite. Plumtree Software involved an offer by an inventor to create a kiosk for a trade show. The inventor created the kiosk using a method he later patented. Interpreting Pfaff, the Federal Circuit in Plumtree Software held that no offer for sale could be made where, unlike here, the inventor agreed to produce a product, but did not necessarily agree to use the later-patented method to produce the product. See id. ("This reference to the software/hardware package is ambiguous as to whether it required MA to provide the kiosk system software or to perform the patented method. Moreover, Plumtree has made no showing that extrinsic evidence would compel an interpretation that MA was bound to perform the patented method."). Thus, the invention at issue in Plumtree Software was a process by which the kiosk was produced. Here, by contrast, there is no claim that MCI required Netscape to use a certain software-writing process to produce the sold product — i.e., the computer source code — implementing the cookies technology.

IV.

Plaintiff next argues that the second Pfaff prong was not satisfied, and thus it was error to conclude that the method disclosed in claim 1 of the `670 patent was "ready for patenting" prior to October 6, 1994, the critical date. Under Pfaff, there are two means of establishing that an invention is "ready for patenting": (i) by adducing "proof that the invention was reduced to practice"; and (ii) by adducing "proof that prior to the critical date the inventor had prepared drawings or other descriptions of the invention that were sufficiently specific to enable a person skilled in the art to practice the invention." Pfaff, 525 U.S. at 67-68. As noted supra, this prong of the Pfaff test must be established by clear and convincing evidence, meaning that "the ultimate factfinder [must have] an abiding conviction that the truth of [the claimants'] factual contentions are highly probable." Pfizer, Inc. v. Apotex, Inc., 480 F.3d 1348, 1360 n. 5 (Fed. Cir. 2007) (citations and quotation marks omitted).

The Memorandum Opinion found by clear and convincing evidence that the cookies invention was ready for patenting prior to October 6, 1994, under both theories. Specifically, the fact that Montulli entered draft cookies source code into the Netscape source code repository on October 3 or 4, 1994, was found to establish, by clear and convincing evidence, that the cookies invention was reduced to practice prior to the critical date. See Netscape II, ___ F. Supp. 2d ___, 1:09cv225, at 25-26. Additionally, the record evidence was also found to establish clearly and convincingly: (i) that Montulli conceived of the invention in July or August 1994; (ii) that Montulli reviewed the design and software architecture of the cookies invention with John Giannadrea — Montulli's supervisor and later Netscape's Web browser division's chief technology officer — during a series of design meetings in July and August 1994; and (iii) that Montulli thereafter began to reduce the invention to computer source code, which he began logging into the Netscape source code repository on October 3 or 4, 1994. Taken together, these facts supported a finding that, prior to the critical date, Montulli made an enabling disclosure that would have allowed Giannandrea, a person skilled in the art, to practice the claimed invention under the Federal Circuit's decision in Robotic Vision Systems v. View Engineering, 249 F.3d 1307, 1312 n. 2 (Fed. Cir. 2001) (holding that the second Pfaff prong is satisfied despite no "actual completion of such software . . ., provided that there is a disclosure that is sufficiently specific to enable a person skilled in the art to write the necessary source code to implement the claimed method").

A.

With respect to the "ready for patenting" Pfaff prong, plaintiff first challenges the conclusion that the cookies invention was reduced to practice solely because Montulli logged draft cookies computer source code into the Netscape source code repository on October 3 or 4, 1994. This challenge succeeds.

Under the Federal Circuit's decision in In re Omeprazole Patent Litigation v. Apotex Corp., 536 F.3d 1361, 1373 (Fed. Cir. 2008), reduction to practice under § 102(b) must be established by proof that the inventor: (i) constructed an embodiment or performed a process that met all the invention's limitations; and (ii) determined that the invention would work for its intended purpose. In some circumstances, "[t]esting is required to demonstrate reduction to practice . . . because without such testing there cannot be sufficient certainty that the invention will work for its intended purpose." Id. (citation and quotation marks omitted).

In this case, an embodiment capable of performing the entire process disclosed in claim 1 would require the existence of both client/browser-side computer source code and server-side computer source code. The client/browser-side source code is necessary to the client computer's storage of the state object, and the server-side source code is necessary to the http server's creation and transmission of the state object in the first instance. Here, Montulli did not construct an embodiment of the claimed invention by writing draft cookies source code by October 6, 1994, the critical date. Although client/browser-side source code was logged into the Netscape source code repository on October 3 or 4, 1994, it is undisputed that server-side source code was not entered into the Netscape source code repository until October 13, 1994, one week after the critical date. See Pl.'s Ex. 42 ¶¶ 22-31 (expert report of Edwin Aoki); Pl.'s Ex. 43 ¶¶ 35-47 (expert report of Judson Valeski). The fact that a client computer may have been able to receive and store the state object is, by itself, insufficient to support a conclusion that the draft browser-side source code constituted an embodiment capable of performing all of the limitations disclosed in claim 1.³⁰

30 Given this, it is unnecessary to reach or address plaintiff's argument that defendants are required to submit expert testimony explaining the content and import of the computer source code. Nonetheless, it is worth noting that the summary judgment record contains a description of the source code by plaintiff's experts that makes sufficiently clear that the source code relates to cookie functionality. See Pl.'s Ex. 42 ¶¶ 22-31 (expert report of Edwin Aoki) (analyzing "cookie specific source code" and identifying "mkaccess.c" and "mkhttp.c" files for cookie functionality relating to Web browsers and http servers); Pl.'s Ex. 43 ¶¶ 35-47 (expert report of Judson Valeski) (same).

Moreover, even assuming that client-side source code was sufficient because, as defendants argue, "[the client — and server-side] software components would necessarily have worked together in a seamless fashion," the record evidence does not establish that the cookies invention was reduced to practice prior to the critical date for two additional reasons. First, Giannandrea's testimony that "by definition, all of the browser features are intended to work with any web server,"³¹ does not prove that Montulli constructed an embodiment that performed the process prior to the critical date. Second, no evidence in the summary judgment record addresses whether the source code as it existed on the critical date, namely October 6, 1994, would have worked for its intended purpose, or whether further testing may have been required to confirm its functionality. See In re Omeprazole, 536 F.3d at 1373.

31 Defs.' Ex. 21, at 70.

In response, defendants make two related arguments concerning whether completed server-side source code is required to prove that the cookies invention was ready for patenting under the reduction to practice theory. First, defendants argue that incomplete source code is sufficient to prove reduction to practice provided that the source code enables a person skilled in the art to practice the invention. This argument fails because it conflates the two separate theories identified in Pfaff under which a party may prove that an invention was ready for patenting. As the Supreme Court explained in Pfaff, the "[ready for patenting] condition may be satisfied in two ways: [1] by proof of reduction to practice . . . or by proof that . . . the inventor had prepared drawings or other descriptions of the invention that were sufficiently specific to enable a person skilled in the art to practice the invention." Pfaff, 525 U.S. at 67-68 (emphasis added). These theories are distinct, and thus even assuming the incomplete source code constituted an enabling disclosure, this fact does not prove that the cookies invention was ready for patenting under the first theory, namely by proof of reduction to practice. Enabling disclosures are relevant only to the second ready for patenting theory. Furthermore, defendants argue that claim 1 does not disclose a particular source code, and therefore the existence of source code is not necessary to prove reduction to practice.³² Although claim 1 describes a particular method and not specific software capable of performing the method, proof of reduction to practice necessarily requires that the process be performed. See In re Omeprazole, 536 F.3d at 1373 (requiring performance of process and determination that invention works for intended purpose). In this respect, there is no doubt that claim 1 can only be reduced to practice and performed where, unlike here, source code fully implementing the claimed method is written prior to the critical date. Cf. Robotic Vision Sys. v. View Eng'g, 112 F.3d 1163, 1166 (Fed. Cir. 1997) ("The patent cannot be held to fail to comply with the best mode requirement for lack of the word `software,' the use of which was plainly apparent to one skilled in the art. Such a disclosure was implicit in the specification.").

32 It is worth noting that in the course of oral argument, defendants' counsel correctly recognized that "[source code] has something to do with reduction to practice." See Transcript at 19 (Mar. 26, 2010).

Accordingly, plaintiff correctly argues that the summary judgment record does not, on this theory, establish that the cookies invention offered for sale was reduced to practice, and thus ready for patenting, prior to the critical date. Yet, the Memorandum Opinion also found that the cookies invention was ready for patenting under a second theory, which is addressed next.

B.

Plaintiff also argues that the alternative ground for concluding that the cookies invention was ready for patenting prior to the critical date — namely that Montulli conceived of the invention and made an enabling disclosure sufficiently specific to allow Giannandrea to practice the claimed invention prior to the critical date — was not proven by clear and convincing evidence. In this regard, plaintiff's arguments, which were plainly raised and rejected in the Memorandum Opinion, remain unpersuasive.

Plaintiff incorrectly reads the Federal Circuit's decision in Robotic in three respects. In that case, an inventor disclosed the invention to a colleague who then wrote computer source code implementing the invention. To begin with, Robotic does not require either Montulli or Giannandrea to state expressly that Montulli provided an "enabling disclosure." Indeed, the Federal Circuit in Robotic found an enabling disclosure in the absence of, and without requiring, any testimonial evidence that the inventor or the person to whom he disclosed the invention subjectively believed that an enabling disclosure had been made. In addition, although plaintiff argues that there is a factual dispute as to whether Giannandrea is a person of ordinary skill in the art, the Federal Circuit in Robotic refers only to "a person skilled in the art." Robotic, 249 F.3d at 1310, 1311, 1312 n. 2, 1313; accord Pfaff, 525 U.S. at 67 (requiring that enabling disclosure be made to "a person skilled in the art"). The Robotic decision does not contain the word "ordinary," and thus does not require that the disclosure be made to one of ordinary skill in the art. Instead, the enabling disclosure must simply be made to one skilled in the art, and in this respect there is no doubt that Giannandrea — a senior software engineer capable of reviewing software architecture³³ — is a person skilled in the art.

33 See Defs.' Ex. 21, at 16, 61 (Giannandrea testifying that he "had more years of experience as a software engineer than [Montulli]," the inventor).

Nor is plaintiff correct in arguing that an enabling disclosure can only be proven under Robotic where the person to whom the disclosure is made actually reduces the invention to practice. Ample authority supports the conclusion that an enabling disclosure can be made even though, as here, the inventor, rather than the person to whom the disclosure is made, ultimately writes the computer source code implementing the invention. Under Pfaff and Robotic, it is the fact of the enabling disclosure itself, not a subsequent reduction to practice based on this disclosure, that satisfies the "ready for patenting" prong of the Pfaff test. As the Federal Circuit explained in clarifying its pre- Pfaff Robotic decision,

[i]n Robotic II, we indicated that, unless the software was completed before the critical date, the method itself could not have been on sale. Robotic II at 1167. However, under Pfaff, actual completion of such software is not required, provided that there is a disclosure that is sufficiently specific to enable a person skilled in the art to write the necessary source code to implement the claimed method.

Robotic, 249 F.3d at 1312 n. 2. Indeed, the facts and holding of the Pfaff decision clearly illustrate this principle. In that case, Pfaff sent detailed drawings of his computer chip socket invention to Texas Instruments one or two months prior to the critical date. Consistent with its practice, Texas Instruments did not immediately construct a prototype for testing and did not begin producing the invention until three months after the critical date. On appeal, the Supreme Court held that proof of the invention's complete conception and the fact of an enabling disclosure satisfied the "ready for patenting" prong of the on-sale bar analysis, and noted that while "[i]t is true that reduction to practice ordinarily provides the best evidence that an invention is complete . . ., it does not follow that proof of reduction to practice is necessary in every case." Pfaff, 525 U.S. at 66. To be sure, as the Supreme Court's statement intimates, evidence of an enabling disclosure is strongest when the enabling disclosure leads the non-inventor to write the computer source code, as in Robotic. Yet, no decision holds that the eventual reduction to practice provides the only means of establishing that a disclosure is an enabling disclosure.³⁴ To announce such a rule here would be to conflate the alternative "reduction to practice" and "enabling disclosure" theories of proving that an invention is ready for patenting under Pfaff, thereby reducing the entire analysis to the single question whether the invention was at some point reduced to practice. Instead, Pfaff and Robotic point persuasively to the conclusion that an enabling disclosure may be found despite the fact that the person to whom the disclosure is made does not reduce the invention to practice. Specifically, where, as here, undisputed clear and convincing evidence establishes (i) that a supervisor, who had more years of programing experience than the inventor, reviewed and assisted the inventor in refining the design and software architecture of the invention, and (ii) that the inventor reduced the invention to practice following these discussions, an enabling disclosure is proven.

34 Plaintiff argues that Space Systems/Loral, Inc. v. Lockheed Martin Corp., 271 F.3d 1076, 1078 (Fed. Cir. 2001), stands for the proposition that an inventor's sending "rough drawings" that "included the system's four steps that are set forth in the claim" does not prove an enabling disclosure. Yet, this argument misreads the decision. The Federal Circuit reversed the decision below because the district court erroneously held that conception of an invention standing alone, as a matter of law, satisfies the "ready for patenting" Pfaff prong, particularly where the invention was further developed after the disclosure. See id. at 1081 ("The district court erred in ruling that the prebias invention was ready for patenting upon conception as communicated in the engineering proposal."). The Federal Circuit did not address the question whether the rough drawings constitute an enabling disclosure.

In this case, there is little doubt that an enabling disclosure was made to a person skilled in the art even though Montulli, not Giannandrea, ultimately reduced the invention to practice. As Giannandrea explained, "I had more years of experience as a software engineer than [Montulli] had, so I was, in age and seniority, senior to him, and I think he was verifying his design idea with me." Defs.' Ex. 21, at 61. More specifically, Montulli disclosed the cookies invention to Giannandrea in July or August 1994, during which time the two programmers met to discuss and diagram on a whiteboard the design and software architecture — that is, the actual computer source code — that might be used to implement the cookies invention. Indeed, in his deposition Montulli describes the extent of the meetings as covering "the entire specification." Defs.' Opp'n Ex. 16, at 131. There is also no evidence disputing Giannandrea's statement that Montulli implemented the invention after these meetings concluded. See Defs.' Ex. 21, at 59. Although neither Montulli nor Giannandrea expressly state that Giannandrea could have reduced the invention to practice based on these discussions, the fact that Giannandrea reviewed and made suggestions to the claimed method not only at the theoretical level, but at the computer source code level, compels the conclusion that Giannandrea was a person skilled in the art who could have written the source code himself.³⁵

35 In this regard, it is worth noting the Federal Circuit has held that ordinarily "creation of the specific source code is within the skill of the art." Robotic, 112 F.3d at 1166.

Contrary to plaintiff's contention in support of the motion at bar, it is clear from the undisputed record evidence that the enabling disclosure occurred prior to the critical date, and that the limitations disclosed in claim 1 were discussed during the meetings. Specifically, Montulli stated in his deposition that he met with Giannandrea in July or August 1994 to discuss the cookies invention's implementation. See Defs.' Opp'n Ex. 16, at 117. Significantly, this evidence is corroborated by Giannandrea's sworn, written declaration to the PTO, in which he states, that the "series of design meetings [were] held during July-August 1994." Defs.' Ex. 23 ¶ 3.³⁶ Additionally, it is equally clear that Montulli and Giannandrea explicitly discussed the state object limitation. When asked in his deposition to describe the exact idea Montulli brought to the summer 1994 design meetings, Giannandrea stated the following:

36 It is worth noting that this declaration is not submitted in support of the `670 patent; rather, it corresponds to the closely-related 6,134,592 patent ("the `592 patent"). As noted in Netscape I, the original application for the `670 patent contained additional claims that the patent examiner concluded could not be claimed in a single patent. Montulli responded on July 23, 1997, electing to prosecute the invention that included claims 1-8, 12-17, and 27-30 in the `670 patent application. These claims were renumbered and issued as the `670 patent. See Netscape I, ___ F. Supp. 2d ___, 1:09cv225, at 4-5. The claims Montulli elected not to prosecute as part of the `670 patent — namely those relating to a Web user's subscribing to an online information service — were nonetheless prosecuted separately and issued as the `592 patent. Significantly, the `670 patent specification and the `592 patent specification are identical, including the figures and examples. Both state that "the invention relates to client-server computer systems in which a server can send state information to a client and the client stores the state information for later retransmissions back to the server." See, e.g., `670 Patent Specification col. 1 ll. 8-11. Accordingly, Giannandrea's declaration is pertinent to the `670 patent because it makes statements relating generally to the conception and implementation of the cookies invention.

Q: All right. Well, what idea did he bring to you specifically, as you recall? A: As I recall, the idea of having a Web server request with [sic] a browser keep a cookie for returning back to the Web server later. Q: So he brought to you the idea, you said, of having the Web server "keep" the cookie? . . . A: The Web server set a cookie. Q: Mr. Giannandrea, your answer — I'll have the reporter read it back, but it's on the screen here. It says: I recall, the idea of having a Web server request with — with a browser keep a cookie for returning back to the Web server later. A: Yes. Defs.' Ex. 21, at 59-60. This evidence, which is corroborated by Giannandrea's PTO declaration statement that, "[i]n one of a series of design meetings held during July-August, 1994," Montulli "disclosed the invention of the subject patent application,"³⁷ plainly establishes that the limitations of claim 1 were discussed during the meeting. Accordingly, the undisputed record evidence clearly and convincingly satisfies the "ready for patenting" prong of the Pfaff test.

C.

Finally with respect to the "ready for patenting" prong of the Pfaff test, plaintiff argues that summary judgment was improperly entered in favor of defendants because plaintiff was not afforded an adequate opportunity to litigate the existence of an enabling disclosure, and therefore was deprived of due process.³⁸ Specifically, plaintiff complains that defendants first identified the Montulli-Giannandrea discussions as proof of an enabling disclosure in their reply brief, thereby depriving plaintiff of an opportunity respond with argument and supporting evidence. In support of this contention, plaintiff cites Eon-Net LP v. Flagstar Bancorp, 249 F. App'x 189, 193-94 (Fed. Cir. 2007), which recognizes that "a court may not sua sponte grant summary judgment on a particular ground without giving the nonmoving party notice and an opportunity to present evidence and argument in opposition."

38 As it is determined supra that it was error to conclude the invention was reduced to practice and thus ready for patenting prior to the critical date, it is unnecessary to reach or address plaintiff's due process claim with respect to that issue.

Even assuming this principle is apposite here, ³⁹ this argument is unavailing; it is flatly contradicted by the record. In the first instance, the parties filed cross-motions for summary judgment on the § 102(b) statutory on-sale bar invalidity defense and fully litigated the issue. Indeed, the parties dedicated a large portion of their summary judgment briefs to addressing this issue and submitted a voluminous record in support of their respective positions. Moreover, in its motion for reconsideration, plaintiff simply repeats arguments that were raised and briefed on summary judgment. For instance, defendants argued in their opening brief in support of summary judgment that Robotic applied, and plaintiff responded by attempting to distinguish Robotic. See Pl.'s Summ. J. Opp'n Br. at 16-17 ("This is an improper reading of the law, and the present action is distinguishable over Robotic Vision."). Similarly, plaintiff argued in its own opening brief in support of summary judgment that Montulli did not make an enabling disclosure to Giannandrea, to which defendants responded by citing evidence of Montulli and Giannandrea's design meetings. See Pl.'s Summ. J. Br. at 22 (arguing that no evidence proves an enabling disclosure); Defs.' Summ. J. Opp'n Br. 30-32 ("Netscape also incorrectly states that `there is no evidence that Montulli had made any other enabling disclosure before [the critical date]'. . . . Again, Netscape's argument misstates the record and completely ignores key facts." (alteration in original)). Significantly, plaintiff's reply on this issue was limited to a single-sentence footnote. See Pl.'s Summ. J. Reply Br. at 20 n. 30 (noting that Giannandrea's testimony shows only that he made unspecified suggestions, "which is clearly not evidence of an enabling disclosure"). Thus, plaintiff's due process argument plainly fails as plaintiff was afforded ample and appropriate notice and an opportunity to litigate fully on summary judgment whether Montulli made an enabling disclosure to Giannandrea prior to the critical date.

39 Plaintiff argues that "[t]he due process problem is a serious concern" and cites a number of decisions purporting to support its position. Yet, the cited decisions are not persuasive in this regard; they are factually distinguishable and address a due process concern not at issue here, namely when a court orders summary judgment on a claim not raised by the litigants. See Fountain v. Filson, 336 U.S. 681, 683 (1949) ("There was no occasion in the trial court for Mrs. Fountain to dispute the facts material to a claim that a personal obligation existed, since the only claim considered by that court on her motion for summary judgment was the claim that there was a resulting trust."); Eon-Net, 249 F. App'x at 193 (reversing sua sponte grant of summary judgment on issues of claim construction and noninfringement because summary judgment motion was limited to issue of license defense); Cooper v. Ford Motor Co., 748 F.2d 677, 680-81 (Fed. Cir. 1984) (reversing sua sponte grant of summary on invalidity based on prior art reference because the parties discussed and cited the prior art reference only with respect to noninfringement). Here, the parties filed and fully litigated cross-motions for summary judgment on the § 102(b) statutory on-sale bar invalidity defense, and thus the entry of summary judgment in favor of defendants based on an enabling disclosure can hardly be described as a sua sponte act that came as a surprise to plaintiff.

V.

In opposing plaintiff's motion for reconsideration, defendants briefly argue that if claim 1 of the `670 patent is invalid under the § 102(b) statutory on-sale bar, then claims 9, 10, and 14 must likewise be invalid because they "add nothing to claim 1 that was not inherent in the product Netscape offered for sale to MCI." This argument fails for two reasons. First, defendants have not filed a formal motion for reconsideration on this issue, and it would be inappropriate to deem defendants' cursory argument in its opposition brief as such a motion. Furthermore, on the merits, the district court decision cited by defendants relies on a withdrawn and superseded Federal Circuit decision. More precisely, the district court decision in Perfect Web Technologies, Inc. v. InfoUSA, Inc. addressed an alleged infringer's contention that a claimed method and a claimed computer system capable of performing the method were invalid on obviousness and anticipation grounds. The district court first agreed that the claimed method was invalid under either legal theory. From this, the district court then reasoned that the five challenged systems claims were also invalid under the Federal Circuit's decision in In re Comiskey, which held that "[t]he routine addition of modern electronics to an otherwise unpatentable invention typically creates a prima facie case of obviousness."<sup>40</sup> Perfect Web, 2008 U.S. Dist. LEXIS 108392, at *29 (S.D. Fla. Oct. 24, 2008) (quoting In re Comiskey, 499 F.3d 1365, 1380 (Fed. Cir. 2007)). Subsequently, however, the en banc Federal Circuit withdrew the original In re Comiskey decision and ordered the panel to reexamine the case. See In re Comiskey, 554 F.3d 967, 969 (Fed. Cir. 2009). Significantly, the revised panel decision — which expressly supersedes the original panel decision — does not state the obviousness principle quoted above and relied on by defendants.<sup>41</sup> Accordingly, defendants' request that claims 9, 10, and 14 of the `670 patent be invalidated as obvious must be rejected.

40 Notably, finding as a legal matter that claims 9, 10, and 14 of the `670 patent are obvious is inappropriate on this record "because the predicate facts underlying an obviousness analysis are plainly contested." Netscape II, ___ F. Supp. 2d ___, 1:09cv225, at 23-24 (identifying disputes of fact concerning the characteristics of one of ordinary skill in the art and the content of the prior art).

41 Indeed, the revised panel decision found that the challenged patent claimed unpatentable subject matter, and thus the decision neither reached nor decided the obviousness issue. See In re Comisky, 554 F.3d at 973 ("We do not reach the ground relied on by the Board below — that the claims were unpatentable as obvious over Ginter in view of Walker, Perry, and `Arbitration Fee Schedule' — because we conclude that many of the claims are `barred at the threshold by § 101.'" (quoting Diamond v. Diehr, 450 U.S. 175, 188 (1981)).

VI.

Claim 1 of the `670 patent discloses a general, broad method in which a state object containing state information is transferred from an http server to a client computer in conjunction with the transfer of a requested file. As found by the Memorandum Opinion, undisputed record evidence establishes, by clear and convincing evidence, that Netscape offered to sell the method disclosed in claim 1 of the `670 patent to MCI before October 6, 1994. Although it was error to conclude on the summary judgment record that the method disclosed in claim 1 was reduced to practice prior to October 6, 1994, the method was nonetheless ready for patenting by that date under Pfaff because the inventor made an enabling disclosure sufficiently specific to allow one skilled in the art to practice the invention. Accordingly, clear and convincing evidence supports the conclusion that claim 1 is invalid under the § 102(b) statutory on-sale bar.

An appropriate Order will issue following the hearing currently scheduled in this matter for 2:00 p.m., Friday, April 16, 2010.