These appeals raise the issues of validity and infringement of claim 3 of Hincher Patent No. 2,841,056. The patent claims a machine which flares one edge of a cardboard strip that is to be used as the vertical component, or side, of a box-notably an inexpensive heartshaped candy box-whose design makes stability difficult. Flaring what is to be the lower edge increases the surface area of contact, allowing the strip to be more securely glued to the base. The patentee does not claim to have invented the concept of a flared edge, or the process of box manufacture, but merely the flaring machine. In this suit Custom Paper Products Company and Hincher, hereafter Hincher, charged that Atlantic Paper Box Company, hereafter Atlantic, had been using a machine that infringed on his invention. Atlantic denied infringement, and asked for a declaration of invalidity.

Two parallel pressure rollers feed the cardboard strip through the Hincher machine. Immediately after the strip passes the rollers, a fixed transverse depressing member exerts a "crushing force" against one edge. The strip is backed up, or "buttressed," by a guide at the opposite edge, so that the edge subjected to the pressure having, so to speak, no place to go, flares out in both directions. In claims 1 and 2 the depressing member is described as a chisel. Claim 3, however, speaks broadly of "edge crushing means." In Atlantic's machine the cardboard strip is projected from endless belts, clearly the equivalent of rollers, and guided so that one edge comes in contact with a rapidly rotating, multitooth cutter wheel, normally used for slotting. The teeth, or cutting surfaces of the wheel, are positioned at cross angles to the strip. Atlantic contended that this wheel was not within the patent. The district court, 340 F. Supp. 897, found otherwise, and that, if valid, Hincher was infringed. However, it held the patent invalid. Both sides appeal.

Before proceeding to the merits we dispose of a procedural issue raised by Hincher. Because the evidence was all documentary and by depositions and affidavits, he urges us to review the evidence de novo and make our own findings. Some courts have taken this approach. See Nasco Inc. v. Vision-Wrap, Inc., 7 Cir., 1965, 352 F.2d 905, 908; Orvis v. Higgins, 2 Cir., 1950, 180 F.2d 537, 539, cert. denied 340 U.S. 810, 71 S. Ct. 37, 95 L. Ed. 595. Others, while shrinking from a de novo hearing on appeal, have suggested that in this situation something less than the "clearly erroneous" standard of review, F.R.Civ.P. 52(a), is appropriate. See, e. g., Caradelis v. Refineria Panama, 5 Cir., 1967, 384 F.2d 589, 593-594. We have not taken either view. See Texas Co. v. R. O'Brien & Co., 1 Cir., 1957, 242 F.2d 526, 529; Leach v. Crucible Center Co., 1 Cir., 1968, 388 F.2d 176, 179, and do not propose to. While in some instances the fact that the district court saw live witnesses may reinforce its conclusions in the area of credibility, cf. United States v. Oregon State Medical Society, 1952, 343 U.S. 326, 331-332, 72 S. Ct. 690, 96 L. Ed. 978, the basic principle remains the same: if the district court's findings, considering the record as a whole, whether based on live or other types of evidence are reasonably supported, they must stand. Our appellate function does not differ just because the parties expedited the trial by obviating the need of live witnesses. Accord: H. K. Porter v. Goodyear Tire & Rubber Co., 6 Cir., 1971, 437 F.2d 244, 246, cert. denied 404 U.S. 885, 92 S. Ct. 203, 30 L. Ed. 2d 169; Worthen Bank & Trust Co. v. Franklin Life Ins. Co., 8 Cir., 1966, 370 F.2d 97, 99-100; Lundgren v. Freeman, 9 Cir., 1962, 307 F.2d 104, 113-114. See generally, C. A. Wright, Federal Practice (2d ed., 1970), p. 431.

Since on initial examination we thought there was no infringement, and that this would be a speedy disposition, we deal first with that issue. We thought there was no infringement for the same reasons, we suspect, that led the examiner to believe originally there was no invention over Claff, Patent No. 2,682,909. Claff, like Atlantic, used a cutting wheel to apply force to the edge of cardboard material. Atlantic contended that since it was only after Hincher distinguished his invention from Claff's that his patent was allowed, he may not now charge that a similar "cutting" machine falls within his claim. The fallacy, however, as Hincher ultimately made clear to the examiner, is that Claff, by using a disc wheel, applied the force parallel to the plane of the material, while Hincher's chisel, like Atlantic's broad wheel, applies it transversely. Hincher disclaimed nothing to escape Claff, and the invocation of file wrapper estoppel is not well taken.

On the basis of our standard of review we see nothing in Atlantic's contention that its machine did not infringe. Taking the broad claim 3 definition of "crushing means," ante, the court was warranted in accepting the testimony of Hincher's expert that while Atlantic's wheel did tear out some of the cardboard fibers, basically it did not cut, but crushed. In all other respects the Atlantic machine substantially duplicated Hincher's.

The only question of any moment is whether Hincher's machine was inventive. As to this the court concluded, 340 F. Supp. 897, 899,

"[P]laintiff's patent does not disclose a patentable invention because plaintiff has failed to persuade me that the patent is the product of 'more ingenuity and skill than that possessed by an ordinary mechanic acquainted with the business.' Graham v. John Deere Co., 383 U.S. 1, 86 S. Ct. 684, 15 L. Ed. 2d 545 (1966). In so ruling, I have in mind the affidavit of defendant's expert, Edward Goldman, which establishes that all of the claimed elements of plaintiff's patent are disclosed in the prior art."

While this language was perhaps on the cryptic side, we agree with Hincher's assumption, based both on content and the citation of John Deere, the leading case on the subject, that the court was finding the patent invalid for obviousness. 35 U.S.C. Sec. 103. Hincher's criticism is that the court made insufficient "basic factual inquiries," 383 U.S. 1, at 17, 86 S. Ct. 684, and, instead, made only irrelevant ones. As to this latter, we agree that the facts that Atlantic made its machine without knowledge of Hincher, and did so in a relatively short time after it perceived a demand for a flared strip, are not "conclusive," Eagle Iron Works v. McLanahan Corp., 3 Cir., 1970, 429 F.2d 1375, 1382, but on the matter of obviousness these circumstances are not irrelevant. Equally, it is not determinative that all of the elements of the patented machine were disclosed in the prior art. It is, however, important.

While it might have saved argument if the court had elaborated more fully the factual bases of its conclusion of noninvention, we think its intent was clearly that the particular combination effected by Hincher was readily apparent to persons of ordinary skill in the trade. We see no basis for criticizing such a finding. There can be no question but that feeding a strip through two parallel rollers to be processed was noninventive.* There was nothing more to the machine except the buttress, socalled, and a crushing member. A restraining guide is as common as a feedroller. In addition, it is clear from the prior art booklet that formed the basis of certain expert testimony, that the concept of a transverse crushing means was well established. For example, the Farmer patent, No. 1,082,687, uses a knife-like tool to flare the top edges of paper cups; and in Taylor, No. 1,966,469, labelled "Method of Manufacture of Flanged Articles of Paper or Like Material," a tool crushes and shapes the edge of a cylindrical container in order to increase the width of the edge. Hincher offers no convincing reason for faulting the court's conclusion that Hincher's particular application called for no unusual skill. The most that could be said is that Hincher established that his machine was novel. This, of course, is not enough.

Affirmed.