MERRITT, Circuit Judge, concurring. I agree with the Court's opinion as far as it goes; but, on the record now before us, I would go further in limiting the scope of the remand. As the Court explains, the Toner Loading Program is not copyrightable because of the merger and scenes a faire doctrines, and even if it were copyrightable SCC's use of the program in this case appears to fall under the fair use exception. Its purpose, though commercial in nature, was only to sell cartridges that could be used by Lexmark printers rather than to profit by infringing any Lexmark copyright. The fact that the Toner Loading Program is not copyrightable defeats both Lexmark's direct claim to copyright infringement and its DMCA claim based on the Toner Loading Program (because the DMCA only prevents the circumvention of measures that protect copyright-protected works). And I agree that Lexmark's DMCA claim based on the clearly copyright-protected Printer Engine Program fails because the authentication sequence does not, and is not intended to, "effectively control[] access" to the Printer Engine Program.

I write separately to emphasize that our holding should not be limited to the narrow facts surrounding either the Toner Loading Program or the Printer Engine Program. We should make clear that in the future companies like Lexmark cannot use the DMCA in conjunction with copyright law to create monopolies of manufactured goods for themselves just by tweaking the facts of this case: by, for example, creating a Toner Loading Program that is more complex and "creative" than the one here, or by cutting off other access to the Printer Engine Program. The crucial point is that the DMCA forbids anyone from trafficking in any technology that "is primarily designed or produced for the purpose of circumventing a technological measure that effectively controls access to a [protected] work." 17 U.S.C. § 1201(2)(A) (emphasis added). The key question is the "purpose" of the circumvention technology. The microchip in SCC's toner cartridges is intended not to reap any benefit from the Toner Loading Program - SCC's microchip is not designed to measure toner levels - but only for the purpose of making SCC's competing toner cartridges work with printers manufactured by Lexmark.

By contrast, Lexmark would have us read this statute in such a way that any time a manufacturer intentionally circumvents any technological measure and accesses a protected work it necessarily violates the statute regardless of its "purpose." Such a reading would ignore the precise language - "for the purpose of — as well as the main point of the DMCA — to prohibit the pirating of copyright-protected works such as movies, music, and computer programs. If we were to adopt Lexmark's reading of the statute, manufacturers could potentially create monopolies for replacement parts simply by using similar, but more creative, lock-out codes. Automobile manufacturers, for example, could control the entire market of replacement parts for their vehicles by including lock-out chips. Congress did not intend to allow the DMCA to be used offensively in this manner, but rather only sought to reach those who circumvented protective measures "for the purpose" of pirating works protected by the copyright statute. Unless a plaintiff can show that a defendant circumvented protective measures for such a purpose, its claim should not be allowed to go forward. If Lexmark wishes to utilize DMCA protections for (allegedly) copyrightable works, it should not use such works to prevent competing cartridges from working with its printer.

Reading the DMCA in pari materia with the rest of the copyright code supports this interpretation. The DMCA should be used as part of the copyright code as it applies to computer software codes and other digital media. To this extent, the specific "purpose" language of the DMCA modifies the more abstract language of the previous copyright law. As the Court explains, the fair use exception in copyright law explicitly looks to the purpose of the one making the copy in determining whether or not such copying violates the statute, and the DMCA itself contains a reverse engineering exception that also demonstrates Congress' s aim merely to prevent piracy. I agree with the Court that both exceptions apply to SCC' s actions in this case. But we should be wary of shifting the burden to a rival manufacturer to demonstrate that its conduct falls under such an exception in cases where there is no indication that it has any intention of pirating a protected work. See, e.g., Lawrence Lessig, Free Culture 187 (2004) (noting the danger that "in America fair use simply means the right to hire a lawyer to defend your right to create"). A monopolist could enforce its will against a smaller rival simply because the potential cost of extended litigation and discovery where the burden of proof shifts to the defendant is itself a deterrent to innovation and competition. Misreading the statute to shift the burden in this way could allow powerful manufacturers in practice to create monopolies where they are not in principle supported by law. Instead, a better reading of the statute is that it requires plaintiffs as part of their burden of pleading and persuasion to show a purpose to pirate on the part of defendants. Only then need the defendants invoke the statutory exceptions, such as the reverse engineering exception. In this case, even if the Toner Loading Program were protected by copyright, and even if the access to the Printer Engine Program were "effectively" controlled, there has been no showing that SCC circumvented the authentication sequence for the purpose of accessing these programs. Indeed, the proof so far shows that SCC had no interest in those programs other than ensuring that their own cartridges would work with Lexmark's printers.

Finally, this reading of the DMCA is also supported by the provision in the Constitution that grants Congress the power to regulate copyright. Article I, section 8, of the Constitution gives Congress the power to regulate copyright in order to "promote the Progress of Science and useful Arts." U.S. Const, art. I, § 8, cl. 8. Congress gives authors and programmers exclusive rights to their expressive works (for a limited time) so that they will have an incentive to create works that promote progress. Lexmark's reading of the extent of these rights, however, would clearly stifle rather than promote progress. It would allow authors exclusive control over not only their own expression, but also over whatever functional use they can make of that expression in manufactured goods. Giving authors monopolies over manufactured goods as well as over their creative expressions will clearly not "promote the Progress of Science and the useful Arts," but rather would stifle progress by stamping out competition from manufacturers who may be able to design better or less expensive replacement parts like toner cartridges.

For these additional reasons, I concur in the Court's opinion reversing the judgment of the district court. On remand the first question should be whether Lexmark can show the requisite "primary purpose" to pirate a copyrighted work rather than to ensure that their own cartridges work with Lexmark's printer. If not, its case against SCC should be dismissed.

CONCURRING IN PART, DISSENTING IN PART

FEIKENS, District Judge, Concurring in part and Dissenting in part. I begin this opinion by noting that my colleagues and I agree on a number of points regarding this case. We agree that this opinion does not foreclose any outcome on the case in chief; instead, its decisions are limited to the grant of the preliminary injunction below. We also agree that Plaintiff Lexmark failed to demonstrate a likelihood of success on the merits of success on the third count, which deals with the Printer Engine Program (PEP) (although I write to explain my own reasoning as to why that is, I also agree with the reasoning of my colleagues). We agree that the Digital Millennium Copyright Act (DMCA) was not intended by Congress to be used to create a monopoly in the secondary markets for parts or components of products that consumers have already purchased. Finally, we agree on the outcome of the second count, although we come to that conclusion for different reasons.

For the reasons explained below, however, I dissent as to the majority's decision on the first count, regarding the copyrightability and infringement of the Toner Loading Program (TLP).

I. Count One

I feel that the record could support a finding that there was enough original expression in the TLP to qualify it for copyright protection. Second, although I agree that the district court erred in its factual findings supporting the conclusion that the TLP was not functioning as a lock-out code, I feel the record offers support for the proposition that it is possible and practical for competitors to make toner cartridges that function with the printer without copying the Toner Loading Program, and therefore, I would remand that issue to the district court to make a determination in the first instance. Third, although I agree with my colleagues that the district court erred in applying the law of the doctrine of merger and scenes a faire, I would apply the doctrines in this case differently.

A. Whether the TLP Contains Original Expression

My colleagues rely on testimony from Defendant's expert Goldberg, who concluded that external factors dictated the way the TLP was written and that the proposals for alternative ways of writing the TLP offered by Plaintiffs expert Maggs were inefficient and repetitive. My colleagues find both that a lookup table does not "differ meaningfully" from the use of equations and that as a matter of law, the equations that were used in the TLP, alternative equations, and a lookup table are "different ideas or methods of operation altogether". I am at a loss to understand how two things that do not differ meaningfully are altogether different. I would hold that each of these alternatives are different expressions of the same idea (monitoring and reporting on toner levels). I would also find that there is factual support in the record for the proposition that there are a variety of expressions that are practical options given the external factors that constrain the program. Nothing in the record suggests that a look up table could not be expressed in the limited number of bytes or the requisite computer language. In fact, given that sample code was provided for alternatives, in my mind, I think it is entirely possible that the district court could conclude, based on this record, that a TLP using (for instance) a lookup table could be written efficiently and practically. Therefore, I would uphold the district court's decision on originality, or at most, remand the case for the district court to reconsider its decision on originality and make more specific factual findings to support its preliminary injunction.

B. Whether the TLP Is a Lock-Out Code

My colleagues and I agree that the district court's finding of fact #79 is inadequate to support the conclusion the district court drew about the practical possibility or impossibility of using the printer without copying the TLP verbatim.^FN1 (If verbatim copying was necessary, then the TLP essentially acted as a "password" or a lock-out code.) However, while my colleagues see the record as supporting only the finding of impossibility, I find the record decidedly mixed on the issue.

There are three ways to make a chip that achieves compatibility between a re-filled Prebate cartridge and the printers at issue: (1) copy Lexmark's chip (and therefore the TLP) verbatim; (2) replace the TLP with a different program and change the checksum value accordingly; or (3) turn the entire TLP sequence "off" with the flip of a single bit. Here, Defendant chose the first option. My colleagues rely on Dr. Goldberg's testimony that the second method is "computationally impossible", and though I do not agree that the record is unchallenged on this point,^FN2 I agree that the record leans strongly toward impossibility as a practical matter. However, I depart from the opinion of my colleagues sharply when I consider the third option.

As I explain below, I believe the record offers strong evidence that the third option is possible, even without contextual information as to the location of the ”off” switch. Therefore, although my colleagues appear to believe that the checksum operation must conclude successfully for the printer to operate, I do not believe the record supports this finding. I believe there is uncontroverted evidence in the record that it is possible to manufacture a chip that never triggers the checksum sequence, and the printer can still operate - in fact, this is exactly what occurs with the non-Prebate cartridges. Further, I believe that the record offers evidence that it was unnecessary for Defendant to copy the TLP to make the cartridge compatible with the printer, even with no contextual information about the location or existence of the "off switch", and would remand to the district court to make new factual findings and determine that question for itself.

Lexmark offered evidence that the third option was possible. In the testimony of Maggs, when the questioning turned to the third option of turning the entire TLP process off with the flip of a single bit, he said it would take "a few days or a few weeks maybe" of trial and error to find the bit that turned off the TLP process. JA 0953 - 4. When asked about the one-bit off switch, Goldberg only stated that he did not know it existed, because Defendants could not determine what the bit did. JA 1030. However, Goldberg stated previously that Defendant ran a very similar test as part of its attempt at reverse engineering, in which it flipped one of every eight bits in the TLP. JA 0575.

My colleagues argue that Lexmark has introduced no evidence that option three, finding the off switch, was "less arduous" than option two, finding a working checksum. This is simply false. Lexmark's expert testified that it would take a "few days or a few weeks" to reverse engineer the one-bit "off switch" by "brute force," assuming you did not know where the "off switch" was. JA 0954. Given this testimony, I think it is improper for my colleagues to decide that reverse engineering a working chip was impossible, instead of remanding this question to the district court.

However, even without Maggs' testimony, I think it is evident from this record that option three is considerably less arduous than option two. The undisputed evidence in the record about the chip's design offers more than enough evidence for a court to find that option three is vastly easier than option two.

The record states that the off switch works regardless of what the values of all other bits on the chip are. JA 0708. Therefore, the maximum number of tests that must be run to find the off switch can be represented as two times the number of bits that might be the off switch (since there are only two values of each bit, 0 and 1). In contrast, the checksum's operation is relative to the values of other bits on the chip. In other words, in contrast to the off switch, a working checksum value depends on the position of 55 other bytes,^FN3 or 440 other bits. JA 0706. This means that no bit can be tested independently, and instead, the tester must search for a "match" between a particular setting of the TLP bytes and the checksum value. Thus, even though there is clearly more than one way to set all the bits to have the chip function, basic principles of mathematics tell us that the number of tests that must be done to find a working checksum and arrangement of bits rises exponentially^FN4 in comparison to the test for single, independently-functioning off switch. Therefore, I think this record amply supports the proposition that it is much simpler to find the off switch than it is to find a working checksum.

The record with regard to what I describe as the third option, reverse-engineering the off switch for the TLP process, could reasonably support the finding that it is possible to find that switch without contextual information, and hence to reverse-engineer a working chip even without a memory map or other contextual information. The factual determination of whether such reverse engineering is possible in turn decides the factual question of whether or not the TLP is a lock-out code. This factual finding significantly impacts any decision about the use of the merger doctrine as a defense to infringement by Defendant, and therefore the likelihood of success on the merits by Plaintiff. Because of the importance of the question, and the insufficiency of the factual findings made by the district court on this question, I would remand to the district court for it to make that finding in the first instance.

C. Doctrines of Merger and Scenes a Faire

1. Merger

I believe my colleagues offer a good description of the law of merger, and I will not restate that law here. However, there is one aspect of merger my colleagues do not decide that I believe is important to this case: a Circuit split regarding the law of merger.

Simply put, there is some disagreement about whether the merger doctrine acts as a bar to copyrightability, or simply as a defense to particular types of infringement. See Nimmer 13.03 [B] [3]; Mason v. Montgomery Data, Inc., 967 F.2d 135, 138, fn 5 (5th Cir. 1992); Kregos v. Associated Press, 937 F.2d 700, 705 (2nd Cir. 1991). The Sixth Circuit has not previously taken a position on this question. The Second and Ninth Circuits have taken the position that merger operates only as a defense to infringement. Kregos v. Associated Press, 937 F.2d 700, 705 (2d Cir. 1991); Ets-Hokin v. Skyy Spirits, Inc., 225 F.3d 1068, 1082 (9th Cir. 2000). The Fifth Circuit holds that merger determines copyrightability. Mason, 967 F.2dl35, 138, fn 5.

As I read the majority's opinion, my colleagues have declined to take a position on this split, but I would do so, because of the importance of this decision to the DMCA count. As I discuss later in my analysis of Count II, the DMCA only protects works that are protected by Title 17 of the U.S. Code or in which the copyright owner has a right under Title 17. 17 U.S.C. 1201(a)(l)(A); 17 U.S.C. 1201(b)(A). Therefore, if the doctrine of merger is applied at the copyrightability stage, and merger is found to have occurred, Plaintiff will have failed to state a claim on which relief can be granted under the DMCA. However, if the doctrine of merger is applied at the infringement stage, then even if merger is found to have occurred, some analysis of the DMCA's scope of protection of the TLP is in order. Therefore, in my view, it is essential to decide whether merger of a work with a method of operation determines the initial question of copyrightability of the work, or operates only as a defense to infringement.

Although I would exercise judicial economy and limit the holding to the case of merger with a method of operation (which is the question I believe this case presents ),^FN5 for the reasons below, I would find the merger doctrine can operate only as a defense to infringement in that context, and as such has no bearing on the question of copyrightability.^FN6

The Copyright Act denies copyright protection to, among other things, methods of operation. 17 U.S.C. §102(b). However, an otherwise copyrightable text can be used as a method of operation of a computer - for instance, an original, copyrightable poem could be used as a password, or a computer program as a lock-out code. In my view, therefore, it is necessary to know what the potential infringer is doing with the material in order to know if merger has occurred. In other words, if I use my own copyrighted poem as a password or lock-out code, an individual who published the poem as part of a book could not escape a finding of liability for infringement. The rationale for the merger doctrine is that without it, certain ideas or methods of operation would be removed from the public realm because all ways of expressing them would be copyrighted. When a poem or program is used as a lock-out code, it is being used as a step in a method of operation of the thing it is locking. Therefore, to protect a work from copying when it is used as a password would be to prevent the public from using a method of operation.

Under this reasoning, an individual who copied a poem solely to use as a password would not have infringed the copyright, because in that scenario, the alleged infringer would have the defense that the poem has "merged" with a method of operation (the password). By contrast, someone who copied the poem for expressive purposes (for instance, as part of a book of poetry) would not have this defense. For these reasons, I would hold that in cases where the merger is with a method of operation, the merger doctrine should be applied as a defense to infringement only, and not as informing the question of copyrightability of the work itself.

Applying that reasoning to this case, the TLP can remain copyrightable as a computer program, and therefore retain some of the protections of copyright law, even if it is a lock-out code. Defendant can still avoid infringement, however, if it uses the TLP only as a method of operation. For instance, Defendant can only claim this defense to infringement if it uses the TLP to interface with the Lexmark printers at issue, and if it is a necessary method of operation of the machine. It would not be able to assert the same merger defense to infringement if it manufactured its own printers and inserted the copied TLP as part of its own Printer Loading Program, even if it originally copied the TLP for the purpose of interfacing with Lexmark printers. Importantly, as this case is concerned, Defendant could not successfully assert this defense if (as a practical matter and not as a theoretical question) the printer could be operated without copying the TLP. In my view, any finding of merger with a method of operation in this case would act as a defense to infringement, but leave the TLP potentially copyrightable (and therefore allow it to qualify for the additional protections of the DMCA).

I support a remand on the question of whether or not the TLP is operating as a lock-out code on the chips in question. Therefore, I would also remand the question of whether the TLP has merged with a method of operation in Defendant's SMARTEK chip, because the two questions are essentially equivalent.

2. Scènes à Faire

In computer programs, the scenes a faire doctrine has been restated as a determination of whether external constraints on the size and content of a program, and standard programming practices, are so deterministic of the content of program that the contents of the program are analogous to the "stock scenes" of literature. Gates Rubber Co.v.Bando Chem. Indus., Ltd, 9 F.3d 823, 838 (10thCir, 1993). Courts have recognized that a rare work may be entirely made up of scènes à faire, but many works will be a combination of scènes à faire and more original elements. See, e.g., Reed-Union Corp. v. Turtle Wax, Inc., 77 F.3d 909, 914 (7th Cir. 1996).

As I read it, Defendant's argument is essentially that because the equations involved in the TLP are standard and their arrangement involved no creativity, the program was one of those rare works that is entirely made up of scènes à faire and the arrangement of those non-copyrightable elements is not creative enough to allow the work any protections of copyright. The record is undisputed on the point that there are elements of the program (some code that translates to LMX) that do not play a functional role and appear to be installed for the purpose of proving verbatim copying in an action like this one. I think it is unquestionable that these elements are not scènes à faire, and they were copied. Thus, I believe that regardless of any other finding, there is a likelihood that some portion of the work would not be classified scènes à faire. However, because it is not clear whether that portion alone would be creative enough to merit copyright, I will discuss the scènes à faire doctrine's application to this case, as I see it.

I believe Feist stands for the proposition that a creative enough arrangement of otherwise uncopyrightable elements would in fact be copyrightable. Feist Publications, Inc. v. Rural Telephone Service Co., 499 U.S. 340 (1991). In that situation, a wholesale copying of the program (such as we have in this case) might still constitute infringement, even though the program is entirely made up of scènes à faire, because the copyrightable arrangement would be copied. In addition, a program (like the one before us) may be a blend of scènes à faire and material that is not dictated by norms or constraints, and in my mind, copying either a substantial part of an original arrangement or copying enough of the non-scènes à faire elements would still constitute infringement.

Finally, there is a Circuit split regarding the stage at which to apply the scènes à faire doctrine that parallels the split as to the merger doctrine. 4 Nimmer§ 13.03[B][4], at 13-76 - 7 n.180.1; Reed-Union Corp. v. Turtle Wax, Inc., 77 F.3d 909,914 (7th Cir. 1996) (holding the doctrine of scenes a faire is separate from the doctrine regarding the validity of a copyright); Ets-Hokin v. Skyy Spirits, Inc., 225 F.3d 1068,1082 (9th Cir. 2000) (recognizing the split and citing Reed-Union favorably on this question); Taylor Corp. v. Four Seasons Greetings LLC, 315 F.3d 1039,1042-3 (8th Cir. 2003) (finding that the scenes a faire doctrine determined whether substantial similarity existed); Hoehling v. Universal City Studios, Inc., 618 F.2d 972 (2nd Cir. 1980), cert, denied, 449 U.S. 841 (1980) (holding that scenes a faire are not copyrightable). The Sixth Circuit has not explicitly taken a position on this split, although in a case regarding "stock scenes" in a movie decided after this case was argued, the Sixth Circuit held that scènes à faire are "too general to qualify for copyright protection" and applied the doctrine before considering substantial similarity. Stromback v. New Line Cinema, Nos. 02-2387/2388, slip op. at 16 (Sept. 14, 2004). The district court took the position that the scènes à faire doctrine should be applied as a defense to infringement, and therefore, on that position alone, this case would require remand for consideration in light of Stromback. ¶37, at 31. My colleagues take the position that Stromback did, namely that the scènes à faire doctrine determines copyrightability.

Clearly, this panel is bound to follow the decision in Stromback. However, I believe that Stromback can and should be distinguished in the area of computer programming. The doctrine of scènes à faire has been expanded in the area of computing beyond the traditional literary "stock scenes" to include those elements that are dictated by external constraints. Gates Rubber Co. v. Bando Chem. Indus., Ltd., 9 F.3d 823, 838 (10th Cir. 1993). Therefore, in this area, application of the scènes à faire doctrine does require some understanding of how the element is being used: as I read the case law, if an otherwise copyrightable arrangement was dictated by external constraints, it would classified as scènes à faire. However, if external constraints, norms, etc. did not dictate a particular arrangement in a defendant's work, but a defendant copied a particular arrangement anyway, it is not clear to me that the case law would still classify that same arrangement as scènes à faire - instead, the work might qualify for some protection from infringement. Thus, I think when applying the extension of the scènes à faire doctrine for computer programs, the district court's rule is superior because it is necessary to understand the circumstances of the copying in order to know whether or not the scènes à faire doctrine applies. Therefore, I would distinguish Stromback and take the other side of the Circuit split for cases regarding the application of the scènes à faire doctrine, when its extension regarding external constraints for computer programs is at issue.

D. Fair Use

I agree with my colleagues that the question in the first element of the fair use defense is "whether the user stands to profit from exploitation of the copyrighted material without paying the customary price." Harper & Row, Publishers, Inc. v. Nation Enters., 471 U.S. 539,562 (1985) (emphasis added). I part ways with my colleagues, however, in their application of that doctrine here.

Here, we have the curious situation that Defendant says it did not know the TLP existed, while arguing that the use is fair. Thus, while Defendant undoubtedly had the purpose of selling exact copies of the Lexmark chip (which contained the TLP) "for private commercial gain" (which weighs against a fair use defense), it apparently did not have the purpose of selling the TLP itself (although it did so).

Harper & Row does say that the defendant's use in that case "had not merely the incidental effect but the intended purpose of supplanting the copyright holder's commercially valuable right." Id at 562 (emphasis in original). Thus, I believe, my colleagues read the case to say that Plaintiff, in order to prove a likelihood of success, must show that the user intended to benefit from the copying of the TLP, and not simply intended to benefit from the sale of an exact copy of the Lexmark chip that happened to contain the TLP.

Here, I part ways. I think that Defendant had the intended purpose of making money from an exact copy of Plaintiff s chip. I do not think that this prong can or should be transformed into a prong weighing in favor of "fair use" because Plaintiffs alleged violation of copyright protections was done unknowingly.^FN7 The Court in Harper & Row said it could not "ignore [the defendant's] stated purpose of first publication." Id. Likewise, I think we should not ignore that this Defendant knew it was making an exact copy of a computer chip, and it did so in order to benefit commercially from whatever content was on the chip. In other words, Defendant copied that chip for the purpose of gaining whatever commercial value such a chip had. Therefore, this factor is at best neutralized by the fact that Defendant did not know that the commercial value was potentially partly derived from copyright rights, and this factor therefore should not weigh against Lexmark.^FN8

With respect to the fourth factor, which involves the effect Defendant's use had on the value of the copyrighted material, I disagree with my colleagues on the question of whether Plaintiff was required to introduce evidence of the independent market for TLPs, or if evidence that the market for the cartridges was impacted was enough. I agree that the Supreme Court instructs us that "[a] use that has no demonstrable effect upon the potential market for, or the value of, the copyrighted work need not be prohibited[...]." Sony Corp. of Am. v. Universal City Studios, Inc., 464 U.S. 417,450 (1984). My difference with my colleagues emerges from my belief that the words "or value of suggest that in situations in which the market for the work itself does not necessarily represent the value of the work, the value of the work is what a court should consider. I believe the value of the TLP is represented by the market for toner cartridges, and therefore, it is not necessary for Plaintiff to introduce evidence about the TLP market in order to demonstrate a likelihood of success on this factor. In other words, the potential commercial benefit from copying the TLP accrues because the TLP on the chip at issue adds functionality to the toner cartridge, which in turn may enhance the market for the cartridge itself.

I think the district court could easily find that a toner cartridge that accurately tells a user when the toner is running low is more valuable than one that does not. More specifically, unlike my colleagues, I think the market for SCC's cartridges containing the SMARTEK chip might have been impacted negatively if the printer inaccurately reported toner levels when using an SCC cartridge. Therefore, as I see it, the TLP on the chip increased the value of Defendant's toner cartridge if the TLP Defendant copied is better able to report on toner consumption than a chip without the TLP would be. If no TLP was on the chip, the printer would use the default TLP code in the Printer Engine Program to warn users of low toner levels. However, there is no evidence in the record that speaks to whether the default TLP that comes with the printer would be able to accurately report the amount of toner in a Prebate cartridge, or whether it would misreport the level of toner to the user. Thus, there is no evidence in the record one way or another regarding whether Defendant's cartridge is more valuable because the cartridge contains the TLP than it would be if the cartridge did not.

In conclusion, I think the market for these cartridges may very well be impacted by the ability of the toner cartridges to make an accurate report of toner levels. I do not believe Lexmark needs to introduce evidence showing that an independent market exists for the TLP -1 think the court can assume an impact to the primary market for the re-filled cartridges, provided that the copied TLP lent additional functionality to the cartridge. Therefore, because the benefit (or lack thereof) gained from copying the TLP on the chip instead of using the printer's default TLP is necessary to deciding the fair use question, and because the record is silent on this point, I would remand to the district court to gather the evidence necessary to make this determination and reconsider its findings on fair use in light of that additional evidence.

As discussed under Count I, the DMCA only offers its additional protections to those works that are already protected by Title 17 of the U.S. Code or those works in which the copyright owner has a protected right under Title 17. 17 U.S.C. 1201(a)(2)(A); 17 U.S.C. 1201(b)(A). Because I believe that this record does not allow the elimination of copyrightability on the grounds of originality, merger, or scènes à faire, I believe some analysis of the DMCA claim regarding the TLP is necessary. However, I note that the DMCA explicitly leaves the defenses to copyright infringement, including the fair use doctrine, unaltered. 17 U.S.C. § 1201(c). Therefore, if the district court on remand were to find that the merger, scènes à faire, or fair use doctrine supplied an adequate defense to infringement, given the copying that went on in this case, I do not believe Plaintiff could meet its burden to show likelihood of success under 17 U.S.C. §1201(b), because there would be no "right of a copyright owner" to prevent the TLP's use in this fashion.

I believe Plaintiff also failed to meet its burden under both 1201 (a) and 1201(b), however, because it failed to present evidence that the chip was primarily designed or produced for the purpose of accessing the TLP.

Interestingly, unlike traditional copyright law, there is an element of scienter present in the DMCA: in order to be a violation of the Act, any technology that is marketed must have been "primarily designed or produced for the purpose of circumventing" a technological measure or other protection of a work (or a portion of the work) protected under Title 17. 17U.S.C. 1201(a)(2)(A); 17U.S.C. 1201(b)(A). Here, I think the evidence before the trial court was overwhelming on the point that while Defendant "primarily designed" the chip to circumvent any protections of the Printer Engine Program, the chip was not "primarily designed or produced for the purpose of circumventing protections for the TLP. Therefore, even if the checksum sequence can be categorized as a lock-out code that must be circumvented in order for the TLP on the chip to be used, or the primary authentication sequence is seen as a protection for both the TLP and the Printer Engine Program, because the chip was primarily designed to allow access to the Printer Engine Program^FN9 and not the TLP, I believe Plaintiff has not demonstrated a likelihood of success on this count.

However, I emphasize the narrow bounds of my position. The record before the trial court at the preliminary injunction stage points overwhelmingly to the conclusion that Defendant did not realize the TLP was on the chip. If evidence showed Defendant knew or should have known that there was a program on the chip, and that it was practical for the Defendant to manufacture a chip that did not access (and therefore use) the TLP, in my mind, that would be a different case. Under that scenario, the defendant would know of the protection, be able to achieve the purpose of operating the printer without circumventing that protection, and yet still choose to circumvent those protections. On such a record, I think a court would have to carefully weigh whether such a situation means that Defendant's product was "produced for the purpose of circumventing" protections of the TLP.

My reasoning on this count is based on my belief that consumers did not have an implied license to use the copyrightable TLP beyond the first re-fill of the Prebate cartridge. With the assumption that the shrinkwrap agreement was valid and enforceable (I believe Lexmark can demonstrate a likelihood of success on that question^FN10 ), I would conclude consumers' implied license to use the copyrighted TLP did not extend beyond the first re-fill of the Prebate cartridge. The TLP at issue is not present in the printer at purchase. Instead, the consumer gains access to using it by purchasing the Prebate toner cartridge (which stores the programs in its microchip). The Prebate toner cartridge is only sold under a special shrinkwrap agreement that requires that the cartridge be returned to Lexmark when it is empty and may not be re-filled by others for reuse. Since the consumer is only authorized to use the Prebate cartridge until the toner runs out, it follows that the license also blocks the consumer from using the TLP after that time.

The TLP was part of the cartridge, and therefore, when the consumer was able to use the TLP after the cartridge had been emptied of toner and re-filled by Defendant (in violation of the shrinkwrap agreement), in my mind, he or she gained unauthorized access to the TLP. However, the record currently supports the finding that Defendant did not have the requisite knowledge of the TLP to form an intent to primarily design or produce a chip for the purpose of gaining access to it. Therefore, I concur on the outcome of the second count, although my reasoning would be different from that of my colleagues.

III. Count Three

In contrast to the TLP, I believe the consumer has a right to use the Printer Engine Program for the life of the printer. Because the consumer has this right, there is no right of the copyright owner to prevent the consumer from using the Printer Engine Program, and therefore, Defendants cannot be found to be in violation of the DMCA. Though the words are never used, I think the concept of this license is present in my colleagues opinion. I agree with their reasoning regarding this count, and write on this issue only because I believe it sheds additional light on my reasoning regarding Count Two.

All the Lexmark printers at issue here come with the Printer Engine Program installed. In fact, it would be impossible for the printer to work at all without such a program, just at it would be impossible for the printer to work without an engine. Therefore, when the consumer buys the printer, the consumer must be buying the right to use not just the physical printer components, but also the Printer Engine Program that allows those physical components to produce printed pages. By buying a Lexmark printer, the consumer acquires an implied license to use the Printer Engine Program for the life of that printer.

The DMCA defines "circumventing a technological measure" to mean avoid, bypass, etc., "a technological measure, without the authority of the copyright owner." 17 U.S.C. § 1201 (a)(3)(A) (emphasis mine). Therefore, under the plain meaning of the law, circumventing a technological measure it is only a violation of § 1201 (a) if the device allows consumers access to a work that they are not otherwise permitted to have. Therefore, even if Defendant has circumvented the authentication sequence to gain access to the Printer Engine Program, or designed a chip with that as its main purpose, it has not violated the statute, because it has not given anyone access to the program who did not already have authority from Lexmark to use it. In fact, it would be impossible to use the toner cartridge's chip to gain illegal access to the Lexmark Printer Engine Program, because only consumers with Lexmark printers would use the toner cartridge, and they already own the right to use the Printer Engine Program.

If this language of the statute were not enough, it is clear from the legislative history that Congress did not intend this provision to apply to devices that merely facilitated legitimate access. In the House of Representative' s Commerce Committee's report on the DMCA, it stated explicitly that the aim of § 1201 (b) was to restrict devices used primarily for piracy, and not those that facilitate legal use of products. In discussion of §1201 (b), the Committee stated,"This provision is not aimed at products that are capable of commercially significant non-infringing uses, such as consumer electronics, telecommunications, or computer products [...] used by businesses and consumers for perfectly legitimate purposes." HR Rep. 1 OS-796 (October 8, 1998). The Committee also stated that §1201(b)(l) seeks to prohibit "making or selling the technological means to overcome these protections and thereby facilitate copyright infringement" [emphasis mine]. Id. In its January 1999 report discussing §1201(b), the Commerce Committee again clarified that the measure was intended to outlaw trade in "devices with no substantial non-infringing uses that are expressly intended to facilitate circumvention of technological measures for purposes of gaining access to or making a copy of a work" [emphasis mine]. HR Rep. 105-846 (Jan. 2, 1999). This understanding of §1201(b) was also clear in the Senate discussions of the DMCA. It stated that the prohibitions in § 1201 (a) were made "meaningful" through the provisions of § 1201 (b), which were intended to enforce "the longstanding prohibitions on infringements." S.Rep. 105-190 (May 11,1998). It reiterated that §1201(b) is intended to attack those devices that "facilitate copyright infringement." Id. Because Defendant's chip can only make non-infringing uses of the Lexmark Printer Engine Program, it is clear Congress did not intend the DMCA to apply in this situation.

IV. Conclusion

For the reasons stated above, I dissent from the majority's opinion with regard to Count One. Although I concur with the result in Count Two, I do not agree with the reasoning of my colleagues and offer my own approach. Finally, I concur with both the reasoning and result for Count Three, but give my own additional reasons for why that is so.

1. I also believe the district court made another clear error in the factual findings relating to the TLP, although it was harmless. In finding of fact #40, the district count found that for one of the printers in question, "7 bytes of Toner Loading Program...are used as an input to the SHA-1 [in the T620/622 printer model]." JA 0273. Yet in finding of fact #67, the district court found that "the contents of the Toner Loading Program are irrelevant to the SHA-1 for calculating the MAC." JA 0278. I believe this second finding of fact is inaccurate; the TLP is not irrelevant to the SHA-1 if it is used as an input. However, because the chip's content dictates the input into the SHA-1, the authentication sequence will complete successfully regardless of the TLP content. Therefore, the district court's error is harmless.

2. On cross-examination, Lexmark's expert, Maggs, said "it wouldn't be easy" to figure out where and what the checksum value should be without contextual information. JA 0953. I read this as contradicting Goldberg's statement that it would be "impossible." In addition, Lexmark offered the statement of another manufacturer of toner cartridges, who said it was not impossible to achieve "a 100 percent working solution" for compatibility with the printer (presumably the 100 percent figure includes access to a program that would accurately report when toner was running low), although "it has become a much more difficult endeavor[..J". JA0214.

3. For the larger of the two TLPs at issue here.

4. Or, depending on the method used to calculate the checksum, factorially.

5. In an analogous situation, the Ninth Circuit chose to address the copying of an access code using the fair use doctrine. Sega Enterprises Ltd. v. Accolade, Inc., 977 F.2d 1510 (1993). I think this situation is better understood as a merger with a method of operation, and not as a fair use. However, I will discuss Sega further when I discuss the fair use doctrine.

6. Nimmer finds this is generally the better view of the merger doctrine. § 1303[B][3].

7. In my mind, this is consistent with the understanding that copyright infringement does not have an element of scienter.

8. Moreover, my colleagues argue that SCC used the TLP for "a different purpose, one unrelated to copyright protection. Rather than using the Toner Loading Program to calculate toner levels...". The record is uncontested on the point that it is the copied TLP that calculates the toner levels for the refilled Prebate cartridges, and governs when the "toner low" message will be displayed. Therefore, Defendant's product did use the program for the same purpose as the Plaintiffs did - to monitor and report on toner levels - even if Defendant did not realize that it was doing so. Therefore, Defendant potentially benefitted from the creative energy that Lexmark devoted to writing the program code. Although a commercial use does not necessarily block a finding of a fair use, Campbell v. Acuff-Rose Music Inc., 510 U.S. 569 (1994), I think this type of commercial use is what Congress intended to weigh against a finding of fair use.

9. I discuss why I believe the DMCA does not reach the Printer Engine Program in my discussion of Count III.

10. SCC contends that such shrinkwrap agreements are not enforceable. In support of this, at least one amicus brief cites a 2001 Federal Circuit court decision that held there must be a "meeting of the minds" in order for restrictions in the agreement to be enforceable. Jazz Photo Corp. v. Int 7 Trade Comm., 264 F.3d 1094,1108 (Fed. Cir. 2001), cert, denied, 536 U.S. 950 (2002). Other circuits have upheld the validity of shrinkwrap agreements. See, e.g., ProCD, Inc. v. Zeidenberg, 86 F.3d 1447 (7th Cir. 1996) (holding that terms inside a box of software bind consumers who use the software after an opportunity to read the terms and to reject them by returning the product). Here, the shrinkwrap agreement was clear and the district court could find that it supports the conclusion that there was a meeting of the minds and the agreement is enforceable. To wit: "This all-new cartridge is sold at a special price subject to a restriction that it may be used only once. Following this initial use, you agree to return the empty cartridge only to Lexmark for remanufacturing and recycling. If you don't accept these terms, return the unopened package to your point of purchase. A regular price cartridge without these terms is available." Finally, I note this case is factually different from Hewlett-Packard Co. v. Repeat-O-Type Stencil Mfg. Corp. Inc., 123 F.3d. 1445 (Fed. Cir. 1997), in which the shrinkwrap agreement contained only a warning against refilling, and did not condition the sale on a promise not to refill.