I write separately from my esteemed colleagues to make several points. First, I wish to distance myself from any sense of moral opprobrium that may be directed toward Kaleidescape. In my *728view, its product was clearly not designed or intended to facilitate the theft of intellectual property; nobody buys it for that purpose; and if it has that incidental effect, it is no worse in most respects—and better in others—than an ordinary personal computer with freely available DVD-copying software.
Second, while I agree that the trial court’s incorporation-by-reference rationale cannot be sustained even under the rules traditionally governing such an issue, I do not think those rules should apply where, as here, the parties have expressly and advertently agreed that one of them will be bound by terms that, for reasons of commercial security, will only be disclosed after that party has executed the agreement. Under traditional incorporation-by-reference analysis, such after-disclosed requirements could never be enforced, because they are never known or available to the licensee until after he has entered the contract. Indeed, Kaleidescape neglects this powerful argument in favor of a more surgical approach only because, I surmise, Kaleidescape recognizes that the traditional rules would prove too much by depriving all of the technical restrictions of contractual force. I do not believe we can invalidate all such restrictions without violating our obligation to adapt the law, as far as possible, to existing commercial practices that operate to facilitate fair commerce. For that reason I would hold the traditional rules inapplicable and would articulate guidelines for determining the enforceability of after-disclosed terms of the type at issue here. I would then remand the matter for consideration in light of those guidelines.
I. The Kaleidescape System
As the case comes to us, the only specific restriction that CCA (Copy Control Association) charges Kaleidescape with infringing is what I will call—for convenience, and without prejudging its meaning or effect—the disk-in-machine clause. This appears in the document “CSS General Specifications” (General Specifications) under the general heading “System Architecture,” where it is part of the description of “the decryption/descrambling process” as it applies to “playback” by either “a DVD-Video Player (stand-alone device)” or “a combination of the DVD-Video DVD Drive and the DVD-Video CSS Decryption Module.” According to CCA, this description contemplates that “playback” will entail the retrieval of data from two physical locations on the original DVD—the “hidden Lead-in Area” and the “hidden Sector Header.” The Kaleidescape system does not comply with this requirement, contends CCA, because once a DVD has been imported into the system, its contents may be played back without retrieving any data from the original DVD. This means the content can be displayed (i.e., the movie can be watched) even if the DVD is not physically present. This in turn makes it possible for the system owner to import a movie from a borrowed or rented DVD and then return the original disk to its owner, retaining a persistent copy on his Kaleidescape system.
*729Whether the disk-in-machine clause is binding on Kaleidescape, and whether Kaleidescape’s products comply with that clause, are to me bloodless issues of law. I see no normative valence in this case at all. My colleagues seem troubled by the fact that, as they note early on, the device “allows users to make permanent copies of borrowed or rented DVD’s so that a user could amass a sizeable DVD library without purchasing a single DVD.” (Maj. opn., ante, at pp. 701-702.) That is true, but at least in the context of the disk-in-machine requirement, it strikes me as scarcely relevant. The machine would still violate that clause—assuming it does—even if it could somehow be made to recognize and reject rented or borrowed disks.1 Indeed it would apparently violate that clause, as construed by CCA, even if it ground the original DVD to powder after storing its contents on the system. The disk-in-machine clause is at most a contract term. It is not a moral imperative.
I recognize that CCA also cites a clause in section 1.5 of the General Specifications (“General Security Requirements”), which states, “The DVD-Video Content Scramble System is intended to prevent casual users from the unauthorized copying of copyrighted materials recorded on DVD-Video/Audio Disks.” But on this point I agree with the trial court, and in substance with Kaleidescape’s expert Daniel Harkins, who described this language as “informative” rather than “normative.” In more familiar legal jargon, it is precatory or advisory rather than operative or prescriptive—a recital rather than a covenant or condition. It informs the rest of the contract, furnishing an aid to its construction, but it does not itself prescribe a rule or obligation that the licensee can breach. Even if it is viewed as prescriptive, its vagueness seems to greatly limit its significance. Who is a “casual” user? What is an “unauthorized” copy?
Whatever the meaning of these terms, the Kaleidescape system appears to have no more tendency to permit “casual users” to engage in “unauthorized copying” than the millions of ordinary personal computers (PC’s) already found throughout this country and the world. Those devices—available for less than one-twentieth the cost of the least expensive Kaleidescape system— furnish far more powerful tools, if slightly less convenient ones, for bootlegging copyrighted DVD’s. For less than $400, a would-be content thief can purchase an ordinary personal computer with a DVD-recording drive—a device now included as standard equipment on nearly all PC’s, and available as an add-on component for well under $50. After installing any of a host of *730widely available software applications (some of them free), the purchaser can load a DVD—his own, a friend’s, or a rental service’s—and strip away all content protection while “ripping” its contents to his hard drive at a per-disk storage cost, at current hard drive prices, of well under $1. He can then copy the content—now stripped of protection—to any number of recordable DVD’s, at a cost, again, of well under $1 per copy. These he can hand out to friends, sell at a flea market, or load into a Kaleidescape-like device, including one that fully complies with the disk-in-machine clause as CCA interprets it.
Indeed an elementary Web search will disclose numerous postings by ordinary consumers who profess to have built, or who express the intent to build, a do-it-yourself Kaleidescape-like system for a tiny fraction of what a real Kaleidescape costs. Unlike the Kaleidescape, these systems will not warn the user against copying DVD’s he does not own. They will exact no promise from him not to do so. And, critically, they impose no constraints whatever on the use he makes of the copied content once he has “ripped” it to his system. In this respect the Kaleidescape device affords an important protection that is entirely missing from a home computer, or a media server built around a home computer: The Kaleidescape can make only the one copy it needs to operate. It furnishes no way for the user to further replicate the content of the DVD, and while its protection system might of course be attacked by a hacker, it was expressly designed to make it “more feasible for an attacker to obtain [encrypted] materials directly from the DVD rather than trying to use a Kaleidescape system to acquire them.” As a result, if the user makes a copy of a DVD he does not own, he will make only one, and that copy will reside only on his Kaleidescape system—nowhere else. He cannot make a second copy to take with him on vacation, or to his second home, or to load on the Kaleidescape on his yacht. The owner of a $400 computer, in contrast, can not only keep a permanent copy on his hard drive for convenient retrieval but can bum an unlimited number of stand-alone, unprotected copies to dispose of as he sees fit.
The ability of any home computer user to copy Content-Scramble-System-protected DVD content onto a recordable DVD seems to severely reduce the utility of a measure that Kaleidescape considered to further deter unauthorized copies: physically sequestering the source DVD in some kind of vault, thus preventing its use for any other purpose, so long as its content resides on the Kaleidescape server. So far as I can discern, a user could defeat this measure by the simple expedient of first burning a copy of the DVD on his PC, then loading the copy onto the system, returning the original to its owner. It is tme that this would add some inconvenience to the process. But that is all it would add to the measures already taken by Kaleidescape to protect the content owner’s interests. To achieve that dubious objective, it would inevitably increase the size, cost, and mechanical complexity of the device, and *731hence its vulnerability to failure. Indeed in one sense it would magnify the problem for the content owner, because every time the system was circumvented by the method just described, there would be two additional copies in existence—one on the server’s hard drive and one on the sequestered (perhaps temporarily) duplicate DVD.2
In sum, while I acknowledge the literal truth of my colleagues’ observation that the device “allows users to make permanent copies of borrowed or rented DVD’s” (maj. opn., ante, at pp. 701-702), I see little intrinsic significance in that fact beyond its furnishing the genesis for this lawsuit. The same capability has been notoriously available for years to any owner of a computer. Furthermore, it seems plain to me that this is an incidental effect of the system, not its purpose. I doubt that anyone would be so fiscally imprudent as to buy a Kaleidescape system to save money on movies. I doubt even more that anyone that imprudent could ever accumulate the capital to afford it. At current price levels the cost of a Kaleidescape system appears to be several times the cost of the DVD’s that will fit on it. Internet reports indicate that a recently introduced “mini” system—apparently the least expensive of Kaleidescape’s systems—carries a recommended price of $8,000. In its base configuration the device will store the content of 75 DVD’s. That means that the purchaser is paying nearly $110 per disk stored—just about 10 times the cost of a typical brandnew DVD. By purchasing additional “mini disk cartridges,” the owner can upgrade the system to hold either 225 or 300 DVD’s (reports vary). The price of such an upgrade is not readily apparent, but before even considering that factor the system would still cost at least $27 per disk stored—several times what a typical DVD costs.
Obviously, anyone who buys such a system to save money on DVD’s is weak on math. The intended purpose of the system has nothing to do with facilitating the unlawful or unsavory acquisition of material the user is not entitled to have and everything to do with exploiting the convenience of data stored in bulk in a digital form. The essential function of the product is to put the owner’s movie collection at his fingertips, as he has long been able to do with his music. When he decides to watch a movie, he need no longer leave his easy chair to pore over shelves, or dig through piles, of disks. He need no longer cross to his entertainment system, place the disk in a DVD player, return to his chair, and wait for the disk to load. He need only sit down before his television and browse electronically through his titles until he finds one he wants to watch. Further, he will never have to wonder again where he, or one of his guests or children, mislaid the family copy of Gone with the Wind (Selznick International Pictures 1939). All his movies are securely *732stored on a hard drive, waiting to be summoned at the touch of a button. The system also offers the capability of viewing his entire collection all over the house—wherever a satellite “player” is connected to the “server” by a secure Ethernet cable—which means that guests or family members can select and view the movies of their own choice, also at the touch of a button. That, not content theft, is the purpose and selling appeal of the system. It is a high-end system in the category as to which the proverb says, “If you have to ask how much it costs, you can’t afford it.”
If Kaleidescape had refused to implement some foolproof copy-protection scheme I might feel differently. But there could be no such refbsal, because there is no such scheme.3 Therefore I cannot condemn Kaleidescape for marketing a system that, as an incident of its core function, stores a copy of a DVD’s content on a secure hard drive.
II. Incorporation by Reference
A. The Trial Court’s Ruling
The trial court ruled that the disk-in-machine requirement was not part of the parties’ agreement because the document in which it appeared—the “General Specifications”—was not explicitly identified in the license agreement as one of the documents it incorporated by reference. I do not believe this ruling is entitled to appellate deference, because it appears to rest on the documents viewed within their four comers. Thus, in a friendly way I disagree with the trial court’s factual premise. Although there is some discrepancy between the description of the binding documents in the license agreement and their actual appearance as supplied to Kaleidescape, that discrepancy is no greater, and in some respects seems slighter, than the *733discrepancies concerning two of the other three documents in question. Kaleidescape concedes that all three of the other documents became part of the contract. I find unconvincing its attempt to carve out the General Specifications for different treatment.
As pertinent here, Kaleidescape’s central obligation under the license agreement was to “comply with the CSS Specifications.”4 “CSS Specifications” is elsewhere defined as “the documentation relating to CSS entitled ‘CSS Specifications’ (including the Procedural Specifications and the Technical Specifications) that Licensor makes available to Licensee, as such documentation may be revised from time to time consistent with [specified provisions].”5 (Italics added.)
Read literally, this definition makes a document part of the “CSS Specifications” with which the licensee must comply if it meets three conditions: (1) it relates to CSS; (2) it is “entitled ‘CSS Specifications’ ”; and (3) CCA “makes [it] available” to Kaleidescape. The reference to “the Procedural Specifications and the Technical Specifications” is explicitly nonexclusive, i.e., the CSS Specifications “include” those documents. By strong implication, they may include others.
It appears that the General Specifications was one of four documents that CCA “made available” to Kaleidescape after the latter signed the license agreement and tendered the license fee. Only one of the four—the Procedural Specifications—may be unreservedly said to conform to the description in the license agreement. At the top of the first page is the legend, “Document Version 2.2 (Effective September 15, 2003) [f] CSS Specifications [f] I. Procedural Specifications.” The other three documents bear the following caption/titles:
DVD-Video Content Scrambling System Authenticator Module for CSS Decryption Module Version 1.10—November 1, 2000 [In larger print:] Authenticator Module for CSS Decryption Module
*734DVD-Video Content Scrambling System: DVD-Video Descrambler Version 1.10—November 1, 2000 [In larger print:] DVD-Video Descrambler
DVD-Video Content Scrambling System: General Specifications Version 1.10—November 1, 2000 [In larger print:] CSS General Specifications
Kaleidescape views the first two of these documents as embodying the “Technical Specifications” explicitly referred to in the definition of “CSS Specifications.” This is a reasonable inference, but they are not so entitled, and nowhere describe themselves as such. Nor do they bear the title “CSS Specifications,” or anything like it. They simply do not conform to the definition of “CSS Specifications” as set forth in the license agreement. That they can be identified as such by inference from circumstances I do not question. But the General Specifications require no resort to circumstances, for on their face they conform substantially, if not unmistakably, to the definition of “CSS Specifications” in the license agreement. There is no question that they were “ma[de] available” to Kaleidescape, or that they “relat[e] to CSS.” And although their title page does not include the precise phrase “CSS Specifications,” it does contain “Content Scrambling System”— the phrase of which “CSS” is an abbreviation—followed by “Specifications,” with only the word “General” in between. I see no reason to doubt that those facts are dispositive of the issue whether they were sufficiently “called out” in the contract.
Kaleidescape misses the point, it seems to me, when it asserts that the license agreement “unambiguously excludes the General Specifications document from the scope of the term ‘Technical Specifications.’ ” (Italics added.) Under the plain terms of the License Agreement, the question is not whether a document falls within the “Technical Specifications,” but whether it falls within the “CSS Specifications” with which the licensee agrees to comply. Kaleidescape’s argument to the contrary rests on the premise that, as it says, “The ‘CSS Specifications’ document has two parts: T. Procedural Specifications’ and ‘II. Technical Specifications.’ ” But this assertion rests on the notion that, by mentioning these particular documents, the License Agreement impliedly excluded other documents that might otherwise conform to the definition of “CSS Specifications.” I cannot accept his implicit invocation of the maxim expressio unius exclusio alterius est, which operates to impliedly exclude unmentioned things when a list of intended things is given. The *735mention of the technical and procedural specifications is explicitly nonexclusive, i.e., the agreement describes the CSS Specifications as “including the Procedural Specifications and the Technical Specifications.” (Italics added.)
In sum, it seems plain to me that the General Specifications is “called out” at least as plainly by the license agreement as two of the three documents by which Kaleidescape concedes it is bound. I would therefore reject Kaleidescape’s incorporation-by-reference argument on its own terms. As will appear, however, I think the legal principles on which the argument rests, though consistent with existing law, cannot be coherently applied in the present context. A different set of principles must be brought to bear if the law of contracts is to fit itself in a rational and just manner to the commercial realities at work when, as here, a technology license is conditioned on technical restrictions that, by express agreement, will not be disclosed until after the licensee has bound itself to comply with them.
B. Commercial Reality
In my view this case presents a fundamental question of contract law: When, if ever, does a contracting party bind himself to covenants or conditions of which, by mutual agreement, he will be made aware only after he enters into the contract? The traditional answer might have been “Never,” because the essence of a contract is an exchange of promises, and it is difficult to conceive of someone effectively promising to act (or refrain from acting) in a manner not then known to him.
Indeed, the traditional conditions for incorporating the terms of another document by reference would categorically forbid enforcement of a promise to be bound by terms that are withheld until after the contract has been executed. Traditionally, for a document to be successfully incorporated by reference, (1) “ ‘ “the reference must be clear and unequivocal,” ’ ” (2) “ ‘ “the reference must be called to the attention of the other party and he must consent thereto,” ’ ” and (3) “ ‘ “the terms of the incorporated document must be known or easily available to the contracting parties.” ’ ” (Shaw v. Regents of University of California (1997) 58 Cal.App.4th 44, 54 [67 Cal.Rptr.2d 850], italics added.) The second requirement essentially ensures that the person agreeing to be bound by an external document recognizes and assents to this aspect of his agreement. That requirement, or its equivalent, must and should apply in every case. But the first and third requirements presuppose that the promisor will either actually know what the external document says, or be provided the means to find out, so that he can be charged with knowledge of its contents. Toward that end, the document must be clearly identified and, if not supplied, at least made available for the promisor’s inspection.
*736Obviously the third requirement would operate to completely defeat any attempt to incorporate a document that is actively withheld from the promisor until after the contract has been made. Neither party finds it advisable to acknowledge this point here, but I find it impossible to ignore, like an elephant in the jurisprudential parlor: A straightforward application of the traditional rules here would mean that Kaleidescape was not contractually bound by any of the four documents it was given after signing the license agreement.6
Kaleidescape’s failure to press such a sweeping argument—though it fairly flies off the page of the authorities it cites—is perhaps not difficult to explain. By invalidating the heart of the parties’ contract—at least under any theory of contract law—the application of traditional incorporation standards would necessarily cast those standards themselves in doubt. Kaleidescape may have hoped to avoid pulling the temple down around its ears, so to speak, by seeking instead to surgically attack the one document that troubled it on the narrow ground that that one document failed to satisfy just one of the traditional requirements. But this approach seems unworkable from a policy perspective, because the requirement on which it relies becomes an empty, pointless formality without the one it ignores. No description of the incorporated document, however precise and accurate, can charge the promisor with knowledge of what he is agreeing to do if the promisor is barred from reading the described document. Once he agrees to be bound by terms to be disclosed later, what good does it do him to describe in precise detail the document where they are written down? Such a description serves its intended purpose only if it permits him to retrieve, recognize, and consult that document before entering the agreement. His ability to recognize it when he sees it is utterly academic if he cannot see it.
Yet application of the traditional rules would unquestionably present great difficulties in some commercial situations where a contracting party may have legitimate business reasons for wishing to withhold certain terms from disclosure until after the other party has bound himself to comply with them. A restaurant franchisor, for example, might quite reasonably balk at disclosing secret recipes to someone who merely professes to be contemplating entry into a franchise agreement. Yet he needs some way to ensure that the franchisee will be bound to follow the recipes, lest the franchise name be tarnished. The franchisor might disclose the recipes, prior to contracting, under a separate confidentiality agreement and assurances of the would-be *737franchisee’s bona fides. But if the parties wish to structure their relationship by binding the franchisee to comply with recipes before they are disclosed to him, and by requiring a franchise fee as a kind of earnest money, we should think twice before overturning their efforts by adhering uncritically to traditional contract law.
For these reasons the parties’ framing of the issue in terms of the clarity with which the license agreement “called out” the General Specifications, and the subordinate arguments about the use of extrinsic evidence and so on to interpret the incorporation clause, strikes me as vaguely off base. If after-disclosed terms of this type are enforceable, then all that can be required prior to their disclosure is that the agreement clearly reflect the promisor’s assent to be bound by those as-yet-unknown terms. Vagueness or imprecision in their description may bear on their enforceability, or at least on the remedies available to the promisee. They may furnish grounds for rescission at the behest of the promisor. But it is difficult to see how they can operate to free the promisor entirely from restrictions to which there is no deeper objection than that the document containing them was misdescribed in the document the promisor actually signed. Here Kaleidescape knew and unambiguously agreed, when it decided to enter into the license agreement, that its use of the licensed technology would be subject to a set of technical restrictions that would only be disclosed if and when it bound itself to comply with them. That, it seems to me, is all that mattered at the contract formation stage. If the conditions did not become part of the contract, it must be for a deeper reason than that the unseen document containing them did not, when finally disclosed, exactly match its description in the signed agreement.7
The real questions are whether courts should enforce such after-disclosed terms at all, and if so under what conditions. I have little difficulty answering the first question in the affirmative. This is the only answer consistent with our obligation, when confronted with novel commercial transactions, to try as *738far as possible to conform the law to evolving commercial custom rather than force commercial actors to conform to the preconceptions of judges. Since at least the days of the Hanseatic League, one of the threads running through the law of contracts has been the desirability of accommodating and facilitating fair and efficient commerce. This thread appears perhaps most famously in the first two avowed purposes of the California Uniform Commercial Code, which are “(1) to simplify, clarify, and modernize the law governing commercial transactions,” and “(2) to permit the continued expansion of commercial practices through custom, usage, and agreement of the parties.” (Cal. U. Com. Code, § 1103, subd. (a)(l)-(2).)
Here the threshold question is whether “custom, usage, and agreement of the parties” militate in favor of the enforcement of technical restrictions such as those embodied, or alleged by CCA to be embodied, in the General Specifications. Certainly the “agreement of the parties” contemplated that Kaleidescape would be bound by a set of technical restrictions even though it would not be told what they were until it had signed the agreement and paid the license fee. The parties have not litigated the questions of custom and usage and it would therefore be rash to opine upon them. However it would hardly be surprising to learn that in license arrangements of this kind, it is common for the licensor to withhold confidential technical restrictions from the would-be licensee until after the licensee binds himself to abide by them. If that is true, I would be extremely reluctant to categorically deprive such provisions of contractual force.
At the same time, by giving effect to such provisions courts will be empowering licensors to sell licensees a pig in a poke, a phrase dating back to an apparently widespread medieval confidence trick. (See Brewer, Dictionary of Phrase and Fable (1898) <http://www.bartleby.com/81/13246.html> [as of Aug. 12, 2009]; Wikipedia, supra, <http://en.wikipedia.org/wiki/ Pig_in_a_poke> [as of Aug. 12, 2009].) If courts are not to become instruments of sharp practice, we must afford some protection against the situation where the pig turns out to be a cat, which scampers off leaving the licensee holding the bag. (See Brewer; Wikipedia.) As safeguards against such situations, the practice should first be limited to circumstances where a reasonable commercial justification was reasonably apparent to the licensor. This suggests that, in some cases at least, a belatedly disclosed restriction may lose some or all of its contractual force because there was no reason it could not be disclosed prior to the promisor’s entry into the agreement. The restriction should also be required to fall within the zone of reasonable expectation from the perspective of the licensee. This criterion might be implicated by an undisclosed restriction unrelated to the subject matter of the agreement, or one that unexpectedly impaired the economic value of the licensed information or technology.
*739These criteria will have to be fine tuned, and others may have to be articulated, as this area of the law evolves in light of commercial practice and judicial experience. In my view, however, they or something like them should determine whether the General Specifications was contractually binding on Kaleidescape. I would remand the matter to the trial court to consider that question in light of these and similar considerations.
Respondent’s petition for review by the Supreme Court was denied October 22, 2009, S176439.
Further, so long as the original disk remains in the Kaleidescape owner’s possession, the fact that he has placed a copy on the system seems no different in substance from leaving a borrowed or rented DVD in an ordinary DVD player. In the case of rented disks, so long as the owner holds onto the original disk he is presumably paying for the content, either as additional rental fees or by foregoing the opportunity to return the disk to the rental service in exchange for a different title.
This last problem might be avoided by another measure Kaleidescape considered, which was to have the device destroy the original disk upon loading its content. I find it hard to imagine that such a system would have any significant commercial potential.
The “cracking” of Content Scramble System (CSS) and the encryption system on which it depends was made possible not by some kind of industrial espionage but by reverse-engineering a licensee’s software product. (DVD Copy Control Assn., Inc. v. Burner (2003) 31 Cal.4th 864, 871-872 [4 Cal.Rptr.3d 69, 75 P.3d 1]; see Wikipedia, The Free Encyclopedia <http://en.wikipedia.org/wiki/Content_Scramble_System> [as of Aug. 12, 2009]; Wikipedia, supra, <http://en.wikipedia.org/wiki/DeCSS> [as of Aug. 12, 2009].) Even without reverse engineering it is possible to “crack” the encryption on a DVD by what is known as a “brute force attack.” The system’s vulnerabilities are summarized on a cooperative online encyclopedia as follows: “The CSS algorithm was soon revealed to be easily susceptible to a brute force attack .... The algorithm’s weakness is primarily due to US government crypto-export regulations, which, at the time, forbade the export of cryptosystems employing keys in excess of 40 bits—a key length that had already been proven to be wholly inadequate in the face of increasing processing power .... In addition, structural flaws in the algorithm reduced the effective key length to only around 16 bits, which could be brute-forced by a 450 MHz processor in less than a minute. As a 450 MHz processor was the stated minimum necessary to decompress a DVD-compliant MPEG-2 videostream in realtime, it effectively meant that any computer that could play a DVD could also crack one.” (Wikipedia, supra, <http://en.wikipedia.org/wiki/Content_Scramble_System> [as of Aug. 12, 2009], fn. omitted.)
“4.2 Compliance with CSS Specifications. [][] 4.2.1 General. Licensee shall comply with the CSS Specifications, as may be amended by Licensor from time to time in accordance with the By-Laws. Each DVD Product shall comply with the version of the CSS Specifications which is in effect at the time such DVD Product is manufactured, taking into account specific effective date provisions in amendments to the CSS Specifications.”
“1.13 ‘CSS Specifications’ shall mean the documentation relating to CSS entitled ‘CSS Specifications’ (including the Procedural Specifications and the Technical Specifications) that Licensor makes available to Licensee, as such documentation may be revised from time to time consistent with Sections 4.2 and 10.7 hereof. Except where otherwise specifically stated, all references to ‘CSS Specifications’ shall be deemed to include all or any portion of the documentation referenced in the preceding sentence.”
Indeed, Kaleidescape agreed to comply with future amendments to the CSS Specifications—an undertaking even more difficult to reconcile with traditional conceptions of contract, and specifically of incorporation by reference. It means that Kaleidescape agreed to be bound by terms that did not exist yet and thus were unknown and unknowable not only to it, but to everyone else, including CCA.
It is of course possible that the promisor in such a case could be prejudiced after the terms are supplied if he is reasonably misled by inaccuracies or ambiguities in the signed contract, or by other deficiencies in the promisee’s communications with him, as to the nature of the materials communicated. Thus the franchisee in my hypothetical case, if sued over some cooking method alleged by the franchisor not to comply with its after-disclosed recipes, might argue that the franchisor was estopped, at least from securing retrospective relief, by the vagueness with which the asserted obligations were communicated. Whether this would entitle the franchisee to continue its allegedly nonconforming practices indefinitely is quite another question. Since the franchisee agreed to comply with the franchisor’s recipes, whatever they might be, a mere delay in understanding what they called for would not seem by itself to justify a permanent exemption from following them. Such an exemption may seem even less defensible in the licensing context, where at least in the realm of federally protected intellectual property interests, licenses are generally interpreted to reserve all rights not expressly granted. (See, e.g., S.O.S., Inc. v. Payday, Inc. (9th Cir. 1989) 886 F.2d 1081, 1088 [“copyright licenses are assumed to prohibit any use not authorized”].)