This is a diversity suit for breach of contract. (It is governed, all agree by Wisconsin contract law, but no peculiarities of that law have been cited to us.) The district court granted summary judgment for the defendants, Tony Richards and his wife. The plaintiff, Dispatch Automation, Inc., has appealed. The particulars of the claims and counterclaims need not detain us; the dispositive issue is the ownership of a computer program called RiMS 2000. If the judge was right that no reasonable jury, on the basis of the evidence gathered in pretrial discovery, could find that the corporation rather than Mr. Richards was the owner, we must affirm.

Richards is a software developer who in 1982 wrote a computer program to help police and fire departments with records management and vehicle dispatch. He called the program RiMS Version 1.0 and continued to develop it, registering copyright on it in 1989, and by 1993 he was up to RiMS Version 5.0. That year he and his wife formed Dispatch Automation with Gary Hagar and his wife, each couple taking a 50 percent interest. The idea was that Hagar, who had experience in marketing, would be the outside partner, and Richards, who would continue developing RiMS, the inside partner. Hagar testified in his deposition that the Richardses needed him because they don't like using the telephone — a sure sign they needed help in marketing.

An attachment to the articles of incorporation defined ownership rights in the corporation's products:

Among the principal products sold by the corporation will be the RIMS group of computer-aided dispatch and records management software products. RIMS is owned by Anthony B. Richards and will be licensed to the corporation. A license fee of $1.00 per year will be paid to Anthony B. Richards. All proceeds from sales of the product will accrue to the corporation. The corporation may continue to develop the product but all ownership rights will remain with Anthony B. Richards.

This was an unusual form of contract in the software industry. Ordinarily an employer insists on owning all the software developed by its employees (unless created wholly on the employee's own time and at his sole expense), whether it is derivative of pre-employment work or completely new, precisely to avoid the kind of dispute that has arisen here. Cf. 17 U.S.C. §§ 101, 201(b). For an employee to own rights to part of the employer's output is bound to create difficult and contentious issues of managing and tracking who owns what, and there is also a danger that the employee will quit and take his technology with him. Then too software development is a risky undertaking and the employer is likely to be the superior risk bearer, and ownership of the software shifts the risk, both upside and downside, from the developer to the firm. But the situation here was unusual. Richards was not an ordinary employee but (with his wife) the half owner of his employer. The company was built around his technology and he was expected to and did continue developing it. The corporation was essentially himself and a marketing team, and he was naturally reluctant to relinquish ownership of the technology that he had invented and would be working to improve. The reservation of ownership in Richards was broadly worded, perhaps precisely to minimize the disputes likely to arise in cases of divided rights.

Successive versions of the RiMS software were developed, culminating in RiMS 2000, also known as RiMS Version 8.0, which was put on the market in 1999. RiMS 2000 was developed by Richards over a two-year period with the aid of an independent programmer for whose services Dispatch Automation paid $46,000.

Around this time the two couples had a falling out, however, and in February 2001 Richards cancelled Dispatch Automation's license to market the "RiMS group of computer-aided dispatch and records management software products" and he and his wife resigned as employees of the corporation. Until the cancellation, Dispatch Automation had been selling RiMS 2000 for roughly eighteen months.

The parties agree that the contract gave Richards no ownership of any new products developed by Dispatch Automation, but only products that were "developments" of the products that he had licensed to the corporation when it was formed in 1993. Dispatch Automation argues that "developments" are small, incremental changes and that RiMS 2000 was so different from the earlier versions that it was a new product. Richards, in contrast, defines a "new" product as one that is not encompassed by the contractual term "RIMS group of computer-aided dispatch and records management software products." Dispatch Automation developed a program for jail management that Richards concedes was not encompassed by the term, presumably because it did not involve vehicle dispatch; another excluded product was a program for the digital imaging of mugshots.

We think that Richards must be right in his understanding of the difference between a new product and the further development of an old one. It would have been cockeyed — it would have been contrary to Dispatch Automation's own interests as they then appeared — for the parties to have agreed that Richards would own successive versions provided they made only incremental improvements over their predecessors but that he would have no rights to a successive version that made a real breakthrough. That would have given him an incentive to pull his punches, or to quit the company if he thought he was on the brink of a breakthrough; neither the articles of incorporation nor, so far as we are aware, any other contractual provision binds Richards to Dispatch Automation. Since the corporation received the entire income (minus $1 a year) from the sale of programs licensed to it by Richards, it had every reason to encourage him to make breakthroughs. Granted, the bigger the breakthrough, the more irksome the 50/50 division of income might seem to Richards. Maybe this was a factor in the falling out of the two couples; but it is hardly to be imagined that Hagar wanted to negotiate a form of contract that would discourage Richards from making his best efforts lest he do so well that he would want the corporate charter revised to give him a bigger slice of the pie. (And it is presumably the absolute rather than relative size of their own slice that would matter to the Hagars.) It is acknowledged that Richards developed RiMS 2000. The assistance of the independent programmer was apparently quite minor; one doesn't buy much time of a first-rate programmer for $46,000, and if he wasn't first rate he probably didn't add much value to the product.

When a contractual interpretation makes no economic sense, that's an admissible and, in the limit, a compelling reason for rejecting it, as we just noted in Hartford Fire Ins. Co. v. St. Paul Surplus Lines Ins. Co., 280 F.3d 744, 747 (7th Cir.2002). "The presumption in commercial contracts is that the parties were trying to accomplish something rational. Common sense is as much a part of contract interpretation as is the dictionary or the arsenal of canons." Fishman v. LaSalle National Bank, 247 F.3d 300, 302 (1st Cir.2001) (citations omitted). And when the senseless interpretation would require indeterminate litigation to implement it, that is another compelling reason for rejecting it. Clark Equipment Co. v. Dial Corp., 25 F.3d 1384, 1387 (7th Cir.1994). "Parties to contracts may prefer simple-minded textualism to costly disputes later on." Hemenway v. Peabody Coal Co., 159 F.3d 255, 258 (7th Cir.1998). They "may prefer, ex ante (that is, when negotiating the contract, and therefore before an interpretive dispute has arisen), to avoid the expense and uncertainty of having a jury resolve a dispute between them, even at the cost of some inflexibility in interpretation." FDIC v. W.R. Grace & Co., 877 F.2d 614, 621 (7th Cir.1989).

To decide, as Dispatch Automation says we (or a jury) must, whether a successive version of a computer software product is a merely incremental improvement or a breakthrough would be like determining the exact moment at which day becomes night. The indeterminacy of that moment is the reason the Weather Service defines "sunset" as the moment when the top of the sun's disk sinks below the horizon from the standpoint of a person standing at ground level. The definition is perfectly arbitrary, but without an arbitrary definition the matter would be indeterminate. And so here. It is doubtful that contracting parties would want to saddle courts with a standard as nebulous as "when the sun sets" to guide decision in the event the parties had a falling out, especially when application of the standard might require a familiarity with computer programming that few judges and jurors have. How for example to decide whether to compare the latest version of a product with the immediately preceding version or with the original version? Compared to the former, it might seem incremental, while compared to the original version it might seem discontinuous, a radical or fundamental change. The first approach would invite the proliferation of meaningless intermediate steps designed to disguise novelty; but the second would ignore the obvious fact that a process of unmistakable, indeed quite gradual, development can have an end point radically different from its beginning, as in the case of the oak and the acorn it grew from. The position for which Dispatch Automation contends is, paradoxically, the minefield that in the usual case leads employers to insist on owning all software developed by their employees.

The primary evidence on which Dispatch Automation relies to show that RiMS 2000 was a new product is a statement that Richards made when the program went on the market that it is "a completely different product underneath, a 100% rewrite of RiMS in a new language with a new database.... RiMS 2000 is four times the size of its predecessor. It does more things, does old things in better, cleaner ways. It takes a bigger PC to run it" (emphasis in original). Even if we ignore the element of sales puffery, we can see that the statement provides an unsatisfactory basis for distinguishing between a new development of an existing product and a new product, though we think it actually leans toward the former, contrary to Dispatch Automation's submission. The reference to the program's being completely new "underneath" and four times larger than its predecessor suggests that the principal changes are in the code rather than the interface with the user, so that unless told that it's completely rewritten and much bigger the user wouldn't know that it had changed. Windows Professional 2000, which many of us judges now use, is the successor to Windows 98, but its code, based on Windows NT, is radically different from that of Windows 98; and yet to many, maybe most, users the only differences between Windows Professional 2000 and Windows 98 is that the new operating system crashes less frequently and is a bit faster. RiMS 2000, with its rewritten and longer code, is said to be both faster and more stable than its predecessors and apparently it's also more user-friendly (fewer steps are required to activate subroutines) and has some new bells and whistles. But from the user's standpoint it is not all that new or different and it continues to fit snugly within the key contractual term: the RiMS group of computer-aided dispatch and records-management software programs.

One might wonder why, if the code for RiMS 2000 is as different from the code for its predecessors as Dispatch Automation claims, neither Dispatch Automation nor Richards has registered it with the Copyright Office. Asked this question at his deposition, Richards answered that "if you tried to submit a copyright application every time there was a new version of the program you would spend all your time trying to continually recopyright the program since ... the program changed literally hundreds of times that day." Registration, of course, is not a condition for copyright protection, though it confers some advantages. See, e.g., Greenberg v. National Geographic Society, 244 F.3d 1267, 1270 n. 3 (11th Cir.2001); Data General Corp. v. Grumman Systems Support Corp., 36 F.3d 1147, 1160 (1st Cir.1994). And contractual alternatives to copyright may give an owner of computer software more protection than copyright would. See, e.g., ProCD, Inc. v. Zeidenberg, 86 F.3d 1447, 1453-55 (7th Cir.1996); Mark A. Lemley, "Beyond Preemption: The Law and Policy of Intellectual Property Licensing," 87 Calif. L. Rev. 111, 124-33 (1999); cf. Oz Shy, "The Economics of Copy Protection in Software and Other Media," in Internet Publishing and Beyond: The Economics of Digital Information and Intellectual Property 97 (Brian Kahin & Hal R. Varian eds.2000).

But all this is by the by, because Richards's avowal of the newness of RiMS 2000 would be relevant only if we adopted Dispatch Automation's interpretation of the contract as requiring that incremental improvements be distinguished from breakthroughs, and we have said that it would not be a proper reading. Dispatch Automation argues that the statement is the kind of objective evidence of "extrinsic ambiguity" that can make a contract which looks clear (that is, contains no "intrinsic ambiguity") unclear. E.g., Rossetto v. Pabst Brewing Co., Inc., 217 F.3d 539, 542-43 (7th Cir.2000); PMC, Inc. v. Sherwin-Williams Co., 151 F.3d 610, 614 (7th Cir.1998); Pierce v. Atchison, Topeka & Santa Fe Ry., 65 F.3d 562, 568 (7th Cir.1995); Bohler-Uddeholm America, Inc. v. Ellwood Group, Inc., 247 F.3d 79, 93, 94 n. 3 (3d Cir.2001); Kerin v. U.S. Postal Service, 116 F.3d 988, 992 n. 2 (2d Cir.1997). We may assume that a written admission by the opposing party is indeed sufficiently objective to create an "extrinsic" ambiguity, cf. FDIC v. W.R. Grace & Co., supra, 877 F.2d at 622, since the requirement of objectivity is intended to prevent the overturning of a straightforward-seeming contractual interpretation on the basis of self-serving testimony by the party challenging that interpretation. Bock v. Computer Associates International, Inc., 257 F.3d 700, 707 (7th Cir.2001); Rossetto v. Pabst Brewing Co., Inc., supra, 217 F.3d at 546; PMC, Inc. v. Sherwin-Williams Co., supra, 151 F.3d at 615; AM International, Inc. v. Graphic Management Associates, Inc., 44 F.3d 572, 575 (7th Cir.1995). (That self-serving testimony might be testimony, admissible at a trial only by virtue of an exception to the hearsay rule, see Fed.R.Evid. 801(d)(2), to an admission by the opposing party; that is why we said that the admission must be in writing to count as objective evidence of extrinsic ambiguity.) But Richards's statement is not that. He was not making any admission about the meaning of the contract. The statement would become relevant if it were decided that the contract requires that incremental and fundamental improvements be distinguished; but that is a step we decline to take.

AFFIRMED.