Harsco Corp. v. Kerkam, Stowell, Kondracki & Clarke, P.C. , 965 F. Supp. 580 ( 1997 )

965 F. Supp. 580 (1997)

HARSCO CORPORATION, Plaintiff,
v.
KERKAM, STOWELL, KONDRACKI & CLARKE, P.C., Edward J. Kondracki, and John C. Kerins, Defendants.

Civil Action No. 1:CV-96-1786.

United States District Court, M.D. Pennsylvania.

May 21, 1997.

Richard G. Tuttle, Linda J. Karpel, Anderson, Kill & Olick, Philadelphia, PA, for Harsco Corp.

C. Andre Washington, Post & Schell, P.C., Philadelphia, PA, for Kerkam, Stowell, Kondracki & Clarke, P.C.

C. Andre Washington, Carol M. Cowhey, Post & Schell, P.C., Philadelphia, PA, for Edward J. Kondracki, John C. Kerins.

MEMORANDUM

CALDWELL, District Judge.

I. Introduction.

In this diversity action controlled by Pennsylvania law, the plaintiff, Harsco Corporation, sued the defendants, Kerkam, Stowell, Kondracki & Clarke, P.C., a law firm, and Edward J. Kondracki and John C. Kerins, two lawyers with the firm. Harsco sets forth two claims. Its first claim is for negligence and alleges that the individual defendants committed malpractice while representing IKG Industries, a division of Harsco, in patent litigation in the Eastern District of Michigan, resulting in a substantial judgment against IKG. The second claim is for intentional or negligent misrepresentation and seeks a declaratory judgment that the defendants improperly told IKG to appeal the trial court's judgment when an appeal would be wholly without merit.

We are considering cross-motions for partial summary judgment filed by the plaintiff and the defendant law firm on the issue of proximate cause in the negligence claim.^[1] The plaintiff argues that, aside from the other elements it must prove to show legal malpractice, the record shows that when the individual defendants moved for judgment as a matter of law under Fed.R.Civ.P. 50(a)(2), their failure to specify the grounds on which they relied was the proximate cause of its injury. Conversely, the defendant maintains that any error (and it denies one was made) did not affect the outcome of the trial.

We will evaluate the motion under the well established standard, see Davis v. Portline Transportes Maritime Internacional, 16 F.3d 532, 536 n. 3 (3d Cir.1994), and we provide the following undisputed background from the summary-judgment record.^[2]

II. Background.

W.S. Molnar Co. (the Company) brought the underlying suit, W.S. Molnar Co. v. IKG Industries, No. 93-CV-60028-AA (E.D.Mich.), aff'd, 82 F.3d 434, 1996 WL 128262 (Fed.Cir.1996) (unpublished disposition), against IKG for infringement of two of its patents.

The individual defendants represented IKG in those proceedings, and one of the defenses they presented was the failure of either patent to disclose the "best mode" of the invention, as required by 35 U.S.C. § 112. A patent that fails to set forth the "best mode" is invalid and cannot be infringed. See Spectra-Physics, Inc. v. Coherent, Inc., 827 F.2d 1524, 1537 (Fed.Cir.1987). Some background on the Molnar patents is necessary to understanding the defense as it was presented at trial.

The patent applications were made on October 20, 1987. The first patent was granted on October 9, 1990, and the second on December 31, 1991. The patents were granted for making articles with a hard, slip resistant surface, primarily for work areas in commercial settings. Part of the process for creating the surface is the heating of a hollow, low-carbon steel wire filled with iron powder, carbon, high-carbon iron particles and aluminum. The invention was claimed to be an improvement over nonslip surfaces that used grit because the patented surface did not wear away. The hardness of the final product depends to a major degree on the carbon content.

The patents advised that:

The total amount of carbon by weight, in the total combination of the hollow wire and fill materials, is selected to be between around 0.25% to around 0.70%. Preferably the amount of carbon is selected to be between around 0.30% to around 0.60% ... The amount of aluminum in the total combination of hollow wire and fill materials is selected to be around 1% to around 3% by weight ... The resultant coating forms a hard, slip resistant surface of iron alloy having a Rockwell hardness in the range from around 40 Rc to around 65 Rc; in a preferred form the result was a slip resistant surface having a file hard surface with a Rockwell hardness in the range from around 50 Rc to around 65 Rc. It should be understood that hardnesses in excess of 65 Rc can be provided.

(Plaintiff's exhibits at A1430 and A1439).

Actually, William S. Molnar, the inventor, preferred a wire that had 2.0% to 4.0% total carbon and 5.0% total aluminum. As the plaintiff points out, this is more than 3 times the carbon content and 1.5 times the aluminum content disclosed in the patents.

The best mode defense focused on the failure of the patents to disclose these percentages of carbon and aluminum. IKG presented evidence that the use of these percentages, by Molnar's own admission, did not give him the hardness he preferred in the surface. Molnar preferred a surface with a Rockwell hardness of 61, and the patent disclosures would have resulted in a hardness of 42. This best mode defense was presented pretrial in a motion for summary judgment (which was denied) and at trial.

To oppose the defense, the Company presented evidence to show that the discrepancy in the figures was unintentional. Molnar testified that he was not a metallurgist, just someone with practical and business exposure to nonslip articles. Additionally, he stated that he did not know the exact composition of the wire because he obtained it from Midwest Thermal Spray which in turn obtained it from a company called Cor-Met. In the months preceding the patent application, Molnar realized that his preferred wire was Cor-Met 1362FC, a wire with the preferred composition of 2.0% to 4.0% total carbon and 5.0% total aluminum, but he did not know its exact composition. Cor-Met supplied the percentages that were listed in the patents when he needed the composition for the patent applications.

As additional support for the innocent nature of the misstatements, his company submitted evidence that, when he wanted to find another supplier for the wire, he used the figures from the patent applications. The other company's wire had to be modified because of the initial inferior results from using the patents' ratio of carbon and aluminum. Molnar also specifically denied that he had intentionally concealed the percentages of carbon and aluminum in the wire.

Further, the Company elicited testimony indicating that a surface in the preferred hardness range could be created from the disclosures in the patents, although not without some experimentation. IKG's vice-president, Mr. Piontek, testified on cross-examination that when IKG prepared its competing product, it simply asked its metal supplier to make a suitable wire. The Company's expert, Mr. Eldis, testified that the patents revealed the "basics" that "would allow you to get to that hardness range" and, as a metallurgist, Eldis could have produced the preferred hardness ranges from the disclosures.

At trial, IKG presented its best mode defense including Molnar's preference for Cor-Met 1362FC at the time of the patent applications in October 1987 and his Company's continued reliance on Cor-Met 1362FC through at least the time of trial. The individual defendants moved at the close of Molnar's case for judgment as a matter of law under Fed.R.Civ.P. 50(a)(1) but, as found by the Federal Circuit on appeal, did not specify the grounds for doing so, as required by Fed.R.Civ.P. 50(a)(2).^[3] The trial court denied the motion and gave the case to the jury. The jury returned a verdict upholding the validity of the Molnar patents.

The individual defendants then filed several post-trial motions, including one that renewed its motion for judgment as a matter of law (JMOL), and specifying that it was relying on a "best mode" defense to the validity of Molnar's patents. It also moved for a new trial. The Company opposed the JMOL motion by arguing that IKG had violated Rule 50(a)(2).

The district court denied IKG's post-trial motions. In doing so, it held that IKG had not properly preserved its right to move for JMOL on the basis of the best mode defense. It also decided that no new trial was warranted because the verdict was not against the great weight of the evidence. The Federal Circuit affirmed. Agreeing that IKG had waived its right to move for JMOL on the best mode defense, the appellate court, although conceding there was "strong evidence of a best mode violation," 1996 WL 128262 at *2, rejected that defense by examining it under a plain error standard of review.

Harsco then filed the instant lawsuit. It alleged that the individual defendants were negligent when they failed to comply with Rule 50(a)(2) by not specifying the facts or the law on which they sought JMOL on the best mode defense, thereby causing the trial court and appellate court not to give this defense the full consideration it deserved. Harsco also alleged that, absent the waiver, it would have prevailed on its best mode defense.

As part of the current summary judgment proceedings, the defendant has submitted other evidence from the trial that bore on other issues. First, it points out that during the application process the patent examiner had objected that the patent was obvious because "routine experimentation" would reveal the "optimum ratio" "depending upon the end results desired" and that "it would be obvious to the practitioner practicing in this art to obtain the optimum ratio of materials utilized."

The defendant also relies on one piece of evidence presented during summary judgment but not at trial. The wire developed for the Company by the other wire manufacturer, Teledyne McKay, although it was the "second or third generation" from the patent specifications, was nonetheless acceptable. (In their opposition brief, the defendant erroneously attributes the development of the wire to IKG.)

III. Discussion.

Under Pennsylvania law:

The three elements of a cause of action for professional negligence (or legal malpractice) are: (1) the employment of the attorney or other basis for his duty to act as an attorney; (2) the failure of the attorney to exercise ordinary skill and knowledge; and (3) that such negligence was the proximate cause of damage to the plaintiff. [citations omitted].

Ibn-Sadiika v. Riester, 380 Pa.Super. 397, 403, 551 A.2d 1112, 1115 (1988) (brackets added).

As noted, both parties move for summary judgment on the element of proximate cause. To establish proximate cause, Harsco has to show that the individual defendants' presentation of the JMOL motion "was a substantial factor in bringing about" its injury. Id. at 406, 551 A.2d at 1116. In other words, the plaintiff has to show that if the motion had been properly presented, Harsco would have prevailed on its best mode defense either at trial or on appeal. Id.

For the purpose of deciding the motions we will also not try to predict how the trial court or the Federal Circuit would have actually resolved the JMOL motion if they had looked at it squarely. Since such a motion presents a question of law, we will simply review it ourselves under the applicable legal principles to see if it would have been granted if there had been no procedural difficulties. See The Charles Reinhart Co. v. Winiemko, 444 Mich. 579, 513 N.W.2d 773 (1994).

Thus, we provide a brief exposition of patent law dealing with the best mode defense and the evidentiary burden IKG had to meet to prevail on this defense in a JMOL motion.

35 U.S.C. § 112 ¶ 1 provides, in relevant part, that the specification "shall set forth the best mode contemplated by the inventor of carrying out his invention." 35 U.S.C. § 112. Determining whether a patent complies with the best mode requirement involves two underlying factual inquiries. First, it must be determined whether, at the time the patent application was filed, the inventor had a best mode of practicing the claimed invention. Chemcast Corp. v. Arco Indus. Corp., 913 F.2d 923, 927-28, 16 USPQ2d 1033, 1036 (Fed. Cir.1990). This inquiry is wholly subjective and addresses whether the inventor must disclose any facts in addition to those sufficient for enablement. Id. at 928, 16 USPQ2d at 1036. Second, if the inventor had a best mode of practicing the claimed invention, it must be determined whether the specification adequately disclosed what the inventor contemplated as the best mode so that those having ordinary skill in the art could practice it. Id. at 928, 16 USPQ2d at 1036-37. The latter question "is largely an objective inquiry that depends upon the scope of the claimed invention and the level of skill in the art." Id. at 928, 16 USPQ2d at 1037.

United States Gypsum Co. v. National Gypsum Co., 74 F.3d 1209, 1213-14 (Fed.Cir. 1996).

Intentional concealment is not required to show a best mode violation. Id. at 1215. Concealment can be either accidental or intentional. Spectra-Physics, Inc. v. Coherent, Inc., 827 F.2d 1524, 1535 (Fed.Cir. 1987). Thus, an inventor cannot defeat a best mode defense by showing that he did not intend to conceal the best mode. Id.

The test for disclosure of the best mode is subjective to the extent it depends on the knowledge of the inventor of a best mode at the time of the patent application, id., and objective to the extent that the disclosure cannot be "so objectively inadequate as to effectively conceal the best mode from the public." United States Gypsum Co., supra, 74 F.3d at 1215 (emphasis in original).

A party challenging a patent on the basis of a best mode violation must prove that defense by clear and convincing evidence. Id. at 1212. Under Fed.R.Civ.P. 50(a)(1), a party moving for JMOL on an issue must show that "there is no legally sufficient evidentiary basis for a reasonable jury to find for that party on that issue." A ruling on a motion for JMOL must take into account the moving party's standard of proof. Thus, for IKG to have prevailed at trial on its motion for JMOL, it had to have shown by clear and convincing evidence that no reasonable jury would have rejected its best mode defense. See AMP, Inc. v. Fujitsu Microelectronics, Inc., 853 F. Supp. 808, 818 (M.D.Pa.1994).

The plaintiff argues that it clearly would have prevailed on its best mode defense. At the time of the patent application, Molnar knew that the best mode of his invention included Cor-Met 1362FC and that this wire was used by his Company in manufacturing its nonslip surface. It was further undisputed that the ranges for carbon and aluminum disclosed in the patents did not accurately reflect the actual percentages in Cor-Met 1362FC, and that the patent disclosures created an inferior product. Moreover, that Molnar may not have intended to conceal the best mode is immaterial since accidental concealment violates the best mode requirement. Hence, the plaintiff argues that the requirement was violated and it should have prevailed on its JMOL motion.

In opposition, the defendant argues the following. First, the trial court's denial of the motion for summary judgment and for new trial indicate that IKG was not entitled to judgment as a matter of law. Under the standards for evaluating these motions, the court necessarily had to find that there was sufficient evidence for a reasonable jury to find for the Company. Hence, the plaintiff cannot claim that the JMOL motion should have been granted. The defendant argues that this position is reinforced when it is considered that the jury instructions, according to the defendant, erroneously imposed a higher standard on the Company to show no best mode violation.

Next, the defendant argues that there was sufficient evidence to defeat the substance of a best mode defense. First, they argue that the evidence showed that Molnar did disclose what he contemplated to be the best mode. Even though he was mistaken, he thought the best mode was represented by the ratios in the patents. Second, the evidence was sufficient to negate a best mode violation since all that is needed to satisfy the best mode requirement is adequate disclosure. Evidence showing that those having ordinary skill in the art could practice the best mode, albeit with some experimentation, satisfied this standard. Third, the wire composition was a mere production detail that did not have to be included in the patent, citing Great Northern Corp. v. Henry Molded Products, Inc., 94 F.3d 1569 (Fed.Cir.1996).

We have considered these and the other arguments of the parties, and we conclude that IKG would not have been entitled to judgment as a matter of law if its JMOL motion had been properly presented at trial.

We turn first to the parties' positions on the substantive elements of the best mode requirement and what a party must show to establish that the patent holder violated it. The plaintiff asserts that the best mode must literally be disclosed, so that the failure to include in the patents the exact composition of the Cor-Met 1362FC wire (or alternatively the trade name or the supplier's name), see United States Gypsum Co., supra, violated the best mode requirement. Conversely, the defendant maintains that the disclosure need only be adequate enough to allow one skilled in the art to make the best mode so that disclosure of the composition of the wire was enough, not the exact percentages of the materials in the composition.

We side with the defendant. As quoted above from United States Gypsum Co., supra, there are two parts to the best mode inquiry. The first part is whether the inventor contemplated a best mode. The second part is whether the specification must have "adequately disclosed what the inventor contemplated as the best mode so that those having ordinary skill in the art could practice it." 74 F.3d at 1214. In United States Gypsum Co., the Federal Circuit reiterated that "the disclosure cannot be so objectively inadequate as to effectively conceal the best mode from the public." 74 F.3d at 1215 (emphasis in original).

Crucially, the court's interpretation of the statutory requirement is that the disclosure be "adequate" or, as it more significantly put it later in its opinion, that the disclosure "cannot be so objectively inadequate as to effectively conceal the best mode." The negative form of the latter phrasing indicates that the disclosure does not have to be precise as long as reveals the best mode in some fashion.

The evidence that would support such a conclusion in the instant case would come from at least two sources: (1) the Company's expert, Mr. Eldis, who stated that the patents revealed the "basics" that "would allow you to get to that hardness range" and that he could have produced the preferred hardness ranges from the disclosures; and (2) IKG's Piontek who stated that his metal supplier manufactured an acceptable wire.

The plaintiff argues that neither of these is proof that a product with the preferred hardness could be produced from the ratio of carbon and aluminum disclosed in the patents. We reject this contention in light of our own conclusion that the best mode requirement deals with an adequate disclosure and only prohibits a disclosure so objectively inadequate as to effectively conceal the best mode. The Eldis and Piontek testimony, even though derided by Harsco as being mere "snippets of testimony," is nonetheless substantive evidence from which a reasonable jury could conclude that one could practice the best mode from the patent disclosures.

The plaintiff objects specifically to the Piontek testimony as showing only that the invention could be reversed engineered. We agree that, if the invention has to be reversed engineered to be understood, then the best mode requirement has been violated, see United States Gypsum Co., supra, 74 F.3d at 1215, but we do not believe that the reverse-engineering objection applies here. Reverse engineering occurs when the inventor's finished product is dissected or taken apart. Id., 74 F.3d at 1215 (rejecting the patent holder's claim that the best mode could be discovered "by examining [the patent holder's] commercial product" since this would require reverse engineering) (brackets added). See also Judin v. United States, 110 F.3d 780, 784 (Fed.Cir.1997) ("nor was any attempt made to dissect" or "reverse-engineer" a sample device); Brooktree Corp. v. Advanced Micro Devices, Inc., 977 F.2d 1555, 1567 (Fed.Cir.1992). In the instant case, on the other hand, the testimony showed that a suitable wire could be developed from using the patent specifications, not from starting with the Cor-Met wire and working backwards.

The plaintiff objects specifically to the Eldis testimony as only showing that a "hardness within the range disclosed in the patents could be produced from the wire composition disclosed in the patents," (Harsco reply brief at p. 11), and that there was "no evidence that Molnar's preferred hardness of Rc 61 could be produced from the composition disclosed in the patents." (Harsco reply brief at p. 11).

This argument is foreclosed by what we said above about the best mode requirement dealing with an adequate disclosure and only prohibiting a disclosure so objectively inadequate as to effectively conceal the best mode. The patents did say the Molnar's preferred hardness range was around 50 Rc to around 65 Rc. This would have gotten one skilled in the art to the preferred hardness of Rc 61.

Our conclusion is buttressed by examining the cases the plaintiff has cited as supporting its claim that the best mode was not revealed: Chemcast Corp. v. Arco Industries Corp., 913 F.2d 923 (Fed.Cir.1990); and United States Gypsum Co., supra. As the defendant notes, both of these cases are clearly distinguishable because in the instant case the patents revealed the chemical composition of the wire and only failed to disclose the ratio of the carbon and aluminum for the preferred mode, a defect a jury could conclude could easily be remedied by one skilled in the art. In the cases the plaintiff cites, on the other hand, an essential disclosure was not made and it prevented practicing the best mode, as a review of the cases illustrates.

Chemcast dealt with the validity of a patent for a grommet, a product used to seal openings in panels and sold to the automobile industry. The best mode inquiry focused on whether the patent had failed to disclose the hardness necessary for the locking portion of the grommet. The patent stated that: "Materials having a durometer hardness reading of 70 Shore A or harder are suitable in this regard." 913 F.2d at 929. The patent failed to disclose the supplier (Reynosol) or the trade name (Reynosol R-4467) of the product used by the patent holder to manufacture the grommet, a product developed after many hundreds of hours. This compound had a hardness of "75 Shore D."

The Federal Circuit held that the best mode requirement had been violated not just because the patent did not "explicitly disclose[]," 913 F.2d at 929 (brackets added), the hardness, the supplier, or the tradename, but because they were also not "implicitly disclosed" to one with skill in the art. Id. Even though the specification mentioned materials of "70 Shore A or harder," it was deficient because materials of 75 Shore D hardnesses were recognized in the art "as different types of materials with different classes of physical properties." Id. at 929. Hence, one skilled in the art could not have practiced the best mode from the disclosures made.

Harsco asserts that Chemcast is on all fours with this case but, significantly, the Federal Circuit did not end its analysis in Chemcast where Harsco insists it should have, after finding no explicit mention of the material hardness, the material itself or the supplier of the material. Instead, the court went on to determine if there had been an implicit disclosure.

We can put aside United States Gypsum Co. for the same reason. That case dealt with the validity of a patent for a lightweight joint compound used to seal joints in wallboard. The best mode used a silicone treated expanded perlite called Sil-42. However, the patent referred only to "expanded perlite ... treated with a silicone compound ..." 74 F.3d at 1211. The Federal Circuit concluded that this did not reveal the best mode because Sil-42 was a qualitatively different material than the various other compounds generally known as silicone treated expanded perlite. Specifically, Sil-42 did not have to be screened before use, resisted breakdown under vacuum treatment, and yielded a lightweight compound that was easy to sand. Id. Thus, because the patent did not "disclose the chemical composition of Sil-42 perlite, its method of manufacture, or supplier trade/name information," id. at 1213, the patent was invalid.

We can easily contrast United States Gypsum Co. with this case. Unlike in that case, the patents here do reveal the materials that composed the wire. Of course, as the plaintiff readily points out, they do not reveal the best ratio of carbon and aluminum to use in the wire. But, unlike in United States Gypsum Co., there is evidence in this case that the optimum ratio was easily determined by one with skill in the art. Indeed, the patent examiner objected to granting the patents on the ground that "routine experimentation" would reveal the "optimum ratio" "depending upon the end results desired" and thus "it would be obvious to the practitioner practicing in this art to obtain the optimum ratio of materials utilized." The plaintiff protests that this comment goes to obviousness rather than to the best mode. Nonetheless, it is still relevant here in light of the Federal Circuit's best mode standard which leaves some play in the second part of the best mode requirement by looking to whether the disclosure was "so objectively inadequate" that it "effectively conceal[s] the best mode." United States Gypsum Co., supra, 74 F.3d at 1215 (brackets added).

Two other cases the plaintiff cites can be similarly distinguished. In Dana Corp. v. IPC Limited Partnership, 860 F.2d 415 (Fed.Cir.1988), the court dealt with the validity of a patent on a valve stem seal sold to the automobile industry. The patent totally failed to disclose a fluoride treatment necessary for satisfactory performance of the seal. The Federal Circuit rejected the patent holder's argument that the best mode had been disclosed because fluoride treatments were well known to those with skill in the art, reasoning that the best mode requirement is not met by a mere reference to the prior art but by what is disclosed in the patent. In our case, there was no total lack of disclosure.

Finally, in Spectra-Physics, Inc. v. Coherent, Inc., 827 F.2d 1524 (Fed.Cir.1987), the court dealt with two patents, one on a laser and the other on a method for constructing a gas laser. The patents revealed that a process known as "TiCuSil brazing" was necessary but did not reveal the essential details of that process as used in the patents. Without those details, the best mode of the invention could not be practiced, even if the person attempting to use the invention knew what TiCuSil brazing was. Again, in our case, because the best mode could be practiced despite the deficiencies in the disclosure, there was no best mode violation.

We are buttressed in our conclusion by the trial court's denial of IKG's motion for new trial. It could have granted that motion despite its denial of the JMOL motion on procedural grounds but declined to do so because it concluded that the verdict was not against the great weight of the evidence. That conclusion means that the trial court believed there was sufficient evidence for a reasonable jury to find for the Company on IKG's best mode defense.^[4] We are also not impressed by Harsco's reliance on the Federal Circuit's statement on appeal that IKG had "strong" evidence of a best mode violation. Strong evidence of a best mode violation can be met with equally strong evidence of no violation, and the Federal Circuit's statement certainly does not indicate that it agrees with Harsco that the Company had no evidence to defeat IKG's best mode defense.

We reject the defendant's remaining defenses. Contrary to the defendant's arguments, Molnar did not disclose what he contemplated to be the best mode. He thought the best mode was represented by Cor-Met 1362FC, not the ratios in the patents. Additionally, the wire composition was not a mere production detail that did not have to be included in the patent. It was an essential part of it. Finally, the jury instructions were not erroneous.

IV. Conclusion.

The plaintiff cannot prevail on an essential element of its legal malpractice claim because it cannot show that the individual defendants' procedural error was the proximate cause of the verdict against it in the underlying patent litigation. This failure is fatal to its claim. Thus, we will enter judgment in favor of the defendant law firm and against plaintiff on the negligence claim.

NOTES

[1] The individual defendants had also joined the law firm's motion, but by order, dated April, 25, 1997, we granted their own motion for summary judgment based on the statute of limitations.

[2] Pursuant to Local Rule 7.9, the plaintiff requests oral argument on the motions. We deny the request because the briefs are adequate to dispose of the motions.

[3] Rule 50(a)(2) provides as follows:

Motions for judgment as a matter of law may be made at any time before submission of the case to the jury. Such a motion shall specify the judgment sought and the law and the facts on which the moving party is entitled to the judgment.

[4] As to the trial court's denial of summary judgment, we place no reliance on that since the court was misled in part by the Federal Circuit's erroneous statement in Brooktree Corp., supra, repudiated in United States Gypsum Co., supra, after the trial court's ruling, that accidental concealment could be a defense to a best-mode violation claim.