Is Article 2B Really Anti-Competitive? ^[1]
Micalyn S. Harris, Winpro Inc.

In a recent article in these pages,^[2] Cem Kaner criticizes the proposed addition to the Uniform Commercial Code of Article 2B - Licensing,^[3] on the ground that "from an anti-competitive viewpoint," the proposal is faulty because it fails to require certain provisions in mass-market licenses which he believes should be mandatory, and fails to prohibit other provisions which he believes are inappropriate.

Mr. Kaner describes himself as "the only advocate for consumers and small businesses who regularly attends" meetings of the Article 2B Drafting Committee, which is a committee of the National Conference of Commissioners on Uniform State Laws "NCCUSL"). By "consumers" he apparently means licensees of mass market software who, because they are individuals, have little "bargaining power" regarding the terms of the mass market license. Under this definition, all of the members of the Drafting Committee, which includes professors, retired lawyers, and practicing lawyers, are "consumers", certainly in their individual capacities, and possibly, in their professional capacities as well. Organizations which use significant amounts of mass market software, including the banking industry and a number of large industrial corporations, are also, under this definition, "consumers", as they purchase mass market software "off the shelf." These organizations are also regularly represented. Even software developers, whom Mr. Kaner does not mention, but whose interests are generally aligned with those he designates as "publishers," are significant users of "off-the-shelf" mass market software, the licenses for which are not negotiable, and thus, developers are also "consumers." The distinction between "consumers" and "publishers" is thus off the mark, if not actually misleading. A more accurate, meaningful, and less emotionally charged distinction is between licensees and licensors.

The NCCUSL Drafting Committee, and most if not all of those who have actively participated in the process, have, as their goal, creating a statute which is clear, even-handed, and will promote commerce in an industry which, over a period of less than two decades, has burgeoned from tiny to in excess of $100 billion.^[4] Of all those involved in the process, it is the small developers who are perhaps most eager to assure that whatever legislation is recommended by NCCUSL is fair and even-handed, because small developers are both licensors and licensees. They must rely on the rights granted in licenses for software they use as well as rely on the licenses they grant to protect their ability to commercialize their software applications. If burdens are to be imposed on developers, distributors and publishers, it is the small developers who are most likely to suffer from the weight of those burdens.^[5]

Mr. Kaner acknowledges that "there is great benefit in creating a uniform legal system for software products and services that works the same way across all states." His objections to the proposed Article 2B are based on his conclusion that the proposed provisions relating to mass market licenses, primarily 2B-208, are anti-competitive. His objections fall into two categories.

The first is a general objection to Section 2B-208's recognition of license agreements as enforceable contracts, generally in accordance with their terms, rather than in accordance with particular terms which he would prefer. The second category focuses on objections to specific license provisions which he claims either should or should not be included in all mass-market software licenses.

Background: Software is a Literary Work

By federal statute, software is a "literary work." As such, copyright inheres in software from the moment of its reduction to a tangible medium (which includes an electronic medium).^[6] Mr. Kaner initially argues that copies of software can (and should) be bought and sold just as copies of books (which are also literary works) are bought and sold. Books, of course, are not sold without restriction. Their "use" (copying or reprinting) is restricted to "fair use".^[7] What constitutes "fair use" of books is not always obvious. Books have a long history and there is statutory and case law on the issue of what constitutes "fair use" of another's copyrighted work.^[8] Nevertheless, what constitutes fair use continues to be the subject of litigation.

By comparison, the mass market software industry is young.^[9] What constitutes fair use of software is much less clear than what constitutes fair use of books.

Although software is, by federal law, a literary work, software is different from books in many ways. "Reading" software, which is done by machine, is different from reading a book, which is done by humans. Humans do not utilize software by "reading" it. Its function is to perform, and in order to perform, software must be copied into a computer's memory. Once copied, the computer "performs" by executing the instructions and manipulating the data which together comprise the software.

Copyright law forbids copying a literary work without permission, except for "fair use," which permits limited copying, as for example, copying limited portions of a book as part of a book review.^[10] Because software must be copied in order to be useful (computers copy software in order to use it), permission to copy must be granted in order to commercialize literary works in the form of software. In some cases, e.g. commercial software code libraries, meaningful commercialization requires that the literary work be copied many times.

Early developers who wished to commercialize their software were, therefore, faced with the challenge of making it possible to permit people to use the software without requiring the developer to give up his copyrights. Developers (whether they undertook to publish and distribute their own software or arranged to have a publisher distribute it for a royalty or on some other basis) were aware that copying software is inexpensive and easy. Reprinting an entire book may well cost more than purchasing a second copy from the author or publisher. Not so software. Making a copy takes little time, and is substantially less trouble and less costly than licensing another copy. Mass market license agreements advise the licensee of the circumstances under which copying is unauthorized, and therefore "unfair" because unauthorized copying violates the licensee's obligations to the licensor.

Developers were also aware (if only after talking with their lawyers) that if they did not make some effort to limit copying, their software might become "public domain." If this occurred, commercial potential would be destroyed.

Developers wanted to be able to provide their creations to a mass market. To do so, developers had to enable their customers to use the applications the developers created, while assuring the developers' ability to obtain income from their creative efforts. In other words, the developers' challenge was to find a way to facilitate intended use of software applications without sacrificing the possibility of commercial gain. The solution which the industry devised was licensing. The use of a contract between developer (whether directly, or through a publisher who would undertake to distribute the software for the developer) made it possible to tailor arrangements to particular applications, and to make changes quickly - an important element in the rapidly developing world of technology.

In essence, licenses tell customers what developers consider "fair use" of their "literary" creations. (Licensing has been used to protect other new technologies. When phonograph records first became available, they were licensed, and the licenses were written on labels on the back of single-sided records. When records had recordings on both sides and labels became smaller, the labels recited references to specific license numbers, e.g., "RCA Victor, licensed under Lic. No.__.")

Early software licenses restricted the licensee to installing the software on a single computer. When Borland amended its license to permit installation on several machines so long as the program was not used by more than one person at a time, it described the revised and extended restriction as permitting use of its software "just like a book."^[11] The relaxed restriction quickly became the standard, but software providers continued to forbid installation on a network server which would make a program available to multiple users simultaneously. Later, program licenses permitted network server installation, but with the proviso that the user would purchase enough licenses to cover the maximum number of simultaneous users served by the server. The beauty of the licensing technique was its flexibility - the ability to change rapidly to adapt to rapidly changing technology and a rapidly evolving market. Licenses were revised quickly to meet changing customer needs, demands and expectations.

The licensing technique also permitted commercialization of code libraries. A code library is typically purchased by a developer for incorporation into applications for third parties. By describing what could and could not be done with a code library, the licensor described what the licensor considered "fair use" of the application. Thus, the developer-licensor could, in effect, say to his developer-licensee, "You may put this into your applications for third parties, but they may not take the code library (either as a whole or its constituent units) out of the application you produce for them and put it into another application. If they want to do that, they have to come back to me and obtain a license, because while you have paid me to use this in your software creations, your customer has not. It's only fair that if your customer wants to use this other than in the application you provide, your customer has to pay me for that privilege."

In such a young and dynamic industry, the use of contracts to advise customers of what software authors see as fair use of their software literary creations has worked well. The industry has expanded, and the variety of software applications today is enormous and growing. Available software ranges from games for young children to systems which run manufacturing operations to applications used by sophisticated developers to create other applications.

The General Objection

Mr. Kaner's general objection may be summarized by his comment that, in essence Article 2B is faulty because "giving publishers the right to create enforceable contracts does not mean that they should be allowed to toss in whatever terms they want, no matter how outrageous."

Article 2B does not give anyone the right to create enforceable contracts containing "whatever terms they want." Article 2B was conceived and drafted as a contract statute. With regard to mass market licenses, 2B-208 provides for enforceability of mass market licenses under basic principles of contract law.^[12] License agreements under 2B-208 are enforceable only to the extent other contracts in our society are enforceable. Terms which are "unconscionable" or against public policy are not enforceable, under Article 2B^[13] or under general principles of contract law.

The theory underlying freedom of contract is that in a market society, supply and demand, the opportunity to obtain commercial benefits, and competition in the marketplace will assure that providers of goods and services will provide what people want and not waste resources on goods and services people do not want.

Where a software license is neither unconscionable nor against public policy, a statute which calls for enforcing it in accordance with its terms reiterates the common law of contracts and is a statement of confidence in the market system, and the ability of the market to reject terms which are unacceptable. In the abstract, one might worry that individual "consumers," that is, licensees of mass market software, have little "bargaining power" in connection with licenses. Experience indicates otherwise. When a license is insufficiently broad to permit a licensee's contemplated use, the would-be licensee either negotiates for what it needs or finds an alternative route to meet its needs. In general, consumers, including but not limited to developers and their distributors and publishers, look to the basic principles of contract law, combined with the pressures of the marketplace, to facilitate commercial exploitation their software creations.

Developers, and their distributors and publishers, understand that unreasonable licenses have effectively "killed" products. For example, several years ago, Borland introduced a new version of its flagship software development product. Customers (in this case, developers) who purchased the new product promptly discovered that Borland had changed its license to provide that a developer would be permitted to distribute up to 10,000 copies of any software developed using the new product, but that for 10,000 copies or more, a separate license would have to be negotiated. Customers realized that if they used the product (as intended) to develop applications which turned out to be very successful, they would be "over a barrel" when they tried to negotiate to sell the 10,000^th copy. They "wrote" to Borland on-line, many returned the product for a refund, and others refused to purchase the product. Within six to 12 weeks, Borland posted a revised license, but a significant segment of the developer community found substitute products or other ways to meet its needs, and Borland lost valuable market share which it never regained.

Thus, licensors are well aware that to the extent they over-reach, they risk losing significant commercial benefits. Evidence to date indicates that the market system works well, and that the current mass market software market is large and highly competitive. To the extent competition is threatened, we have specific laws, i.e., the anti-trust laws, to deal with the problem.

There is no evidence that the market is in need of regulation by way of having a uniform body of law dictating what may and may not be included in software licenses. Mr. Kaner's objections appear to be based more on what he would like included in the price of mass-market software than on what the market as a whole needs in order to be maintained as a competitive market.

Mr. Kaner's second group of objections are to specified provisions in license agreements which, he claims, are typical, have the effect of adversely affecting competition, and if Article 2B is adopted, will be enforced in accordance with their terms. These objections are:

Taking the easy one, Item 2, first, we have seen "beta test" license agreements which prohibit criticizing an application. The reason for including a provision prohibiting criticism of a beta test version is obvious on a moment's thought: publishing criticism of an application made available for testing so it can be improved would not be fair.

Applications are provided for beta test so the developer can find out what needs fixing and fix it. For a beta tester to publish criticism of a beta test application might spoil the market for an application which had considerable potential commercial value. Denying a beta tester the right to criticize the application being tested is essential to preserving the potential market. A beta tester providing his or her criticisms to the world at large, rather than to the developer, would be a breaching contractual obligations of good faith and fair dealing. Stating the provision in the beta test license is no more than a reminder of what is fair use, good faith and fair dealing.

Where such provisions are included in licenses for final versions, they have not been enforced, as is evident from the fact that reviews of new applications are widely available, and often eagerly sought by developers. Newspapers and magazines have regular columns in which new software applications are reviewed and compared with similar applications, or applications intended to produce comparable results. For example, one can find weekly reviews of financial software products in The Wall Street Journal, and monthly reviews in investor magazines such as the American Association of Independent Investors' publication. When they are favorable, copies of reviews are included in promotional materials for an application.

In short, the alleged problem described in Item 2 does not exist.

Items 3 and 4, regarding restrictions on reverse engineering, are really the same. It is true that many licenses include restrictions on reverse engineering, but these restrictions have not been enforced to prohibit interoperability.^[14] Nor is there any evidence that these restrictions have had a "chilling" effect on competition. On the contrary, people have engaged in reverse engineering in connection with creating competitive products, been challenged, and when the restrictions were seen as anti-competitive or otherwise against public policy, courts have refused to enforce them.^[15] Attempts at enforcing prohibitions on reverse engineering to prevent interoperability have been made, and have failed.^[16] Reverse engineering to avoid infringing another's copyright has also been deemed to be "fair use" and therefore, as a matter of public policy, permissible, as has reverse engineering for purposes of understanding, describing and criticizing the functioning of applications.^[17]

As a practical matter, software has become increasingly complex, and in many if not most cases, it is easier and faster for a developer to create a functional equivalent "from scratch" than to reverse engineer an application. When, however, the result of reverse engineering is to reuse code created by another, such reuse is clearly not within the permission intended to be granted by the original license, nor is it necessary to achieve interoperability. Possibly as a result of case law on this issue, many licenses no longer prohibit reverse engineering. Those which do appear unlikely to deter competitive development, but may provide a basis for challenging uses resulting from reverse engineering which the licensor believes are unfair.

Item 1 involves an issue which is, theoretically, more troubling. Under the standard rules of contract formation, the terms of the contract must be disclosed prior to entering into the contract, or the contract is either void ab initio, or voidable. Both the common law and Article 2 of the Uniform Commercial Code have, however, permitted enforcement of contracts originally entered into with open terms which are subsequently "filled in" by the parties. License agreements are generally too long to put on the outside of a software package. The result has been "shrinkwrap" and "clickwrap" licenses. These licenses provide a means for the customer to read the license before opening an envelope or shrinkwrap around the diskette containing the software, and usually provide that if the customer finds the license unsatisfactory, the unopened diskette may be returned for full credit.^[18] The clickwrap provides a means for the customer to read the license before installing the software, and will abort installation if the customer does not agree to the license.^[19] In these cases, if the customer has been billed prior to completion of installation, the customer is also generally permitted to return the software for full credit. In some cases, customers must contact the developer to obtain a return number; but the return of the diskette by mail is usually permitted, and usually a minor inconvenience (as compared with, for example, going to a retail store to return a suit in person).

Even if all of the license terms were included on the outside of a software package, they would be available for review to relatively few customers. This is because distribution from retail stores accounts for only a few of the thousands of software applications made available to the mass market. Shelf-space in retail stores is limited, and reserved for the most popular software packages, most of which are produced by the larger companies. In terms of the number of different software applications, the vast majority is distributed via catalogs, and ordered via mail, fax or telephone. For small developers, the most effective method of distribution is often a targeted mailing and fulfillment via mail, fax and telephone orders.

It is possible for retail stores to make license agreement provisions available to customers. Egghead Software maintained a file of software licenses for every package it stocked, and customers could review these licenses. One had only to ask. Egghead Software, at least as a retail store operation, is now out of business, which may indicate, among other things, that being able to review software licenses was not a significant competitive advantage or otherwise a matter of concern to customers.

In any case, suppose there was a law that before shipping software, a licensor had to provide the customer-licensee with a copy of the license agreement and an opportunity to review it. Would the result benefit anyone? and if so, whom? Large software providers with packages available at retail would not suffer greatly. Retailers could be required to make licenses available to customers at the store. Catalog sales however, would become considerably less efficient than they are now. Software distributed via catalog would become more expensive because prior to shipping the software, a copy of the license agreement would have to be sent out, and an acknowledgment of its acceptability linked up to the ordering customer. The avenue of distribution most available to small developers, publicity through mailing lists and shipment based on orders placed by mail, fax or telephone, would become an administrative nightmare if a copy of the license agreement had to be sent first, and an acknowledgment of its acceptability received before software was shipped. The result would be to reduce competition for mass-market software, increase prices for the software which was available, and make it much more difficult for small developers to distribute software applications, thereby leaving the market to large providers. Such a result is far more anti-competitive than the alleged anti-competitive impact arising from customers' inability to compare license provisions before receiving a shipment of returnable software.^[20]

It is also not clear that customers regard providing the specific terms of a license agreement prior to shipment as information worth paying for. Providing the information, particularly for catalog and direct mail sales, which are the primary methods of distribution available to smaller developers, has a cost which would have to be reflected in the price paid to them. Mr. Kaner may believe that people choose software on the basis of what terms are or are not in a license agreement, but reviews rarely mention this factor as important. There is no evidence that the assumed benefit arising from preshipment review of license agreement terms is commensurate with the additional costs and burdens, which are particularly onerous for the small developer.

Proposed 2B-208 of Article 2B would require licensors distributing through channels which do not permit review of license agreements prior to shipment to accept return of shipped software - not just because the license agreement is unsatisfactory, but for any reason. This is a right which in other industries is granted at the discretion of the distributor, as a marketing technique, and not required by statute. However theoretically troubling the current "shrinkwrap" arrangement may be, permitting software to be shipped and returned unopened (for any reason, including but not limited to an unacceptable license agreement) is, particularly for the small developer, far more desirable than requiring preshipment disclosure, which has a cost that can only make software more expensive for developers to provide and therefore, more expensive to users.

The right of return is a right which customers do not now have, either for software or other items ordered from catalogs, although many catalogs accept returns for marketing reasons. Thus, proposed Article 2B would give customers a right which they do not now have. Mr. Kaner, however, does not find the proposed arrangement satisfactory because if the customer must go to the trouble of returning software, he may not seek an alternative because the customer is no longer "in a shopping frame of mind."^[21]

Destroying or significantly reducing the channels of trade for thousands of software applications in order to permit preshipment review because that is when the customer is "in a shopping frame of mind" seems, as a matter of public policy, a poor choice. Society does not impose such requirements on other items. If one brings home a shirt which turns out to be the wrong color, one may or may not have a right to return it, and in the absence of a defect, will have to bear the time and expense of a permitted return. It is not at all clear why software should be burdened with obligations not imposed on other items. As a matter of public policy, we do not insist that suppliers accept returns, or that they pay the cost of customer returns. Nevertheless, Article 2B as presently drafted, imposes that burden on software licensors if license terms are not available prior to payment and delivery of the licensed software. If there were any public policy concerns regarding inability to review license terms prior to shipment and payment, such concerns should be laid to rest by 2B-208, which requires a licensor who does not make license terms available prior to shipment and collecting payment, to place the licensee in as good a position as if he had reviewed such terms prior to ordering the software and found the terms unacceptable.

That Mr. Kaner's objection is not really to absence of knowledge of license provisions, but to absence of particular services he would like, is revealed when one looks at the "other" anti-competitive effects of which Mr. Kaner complains. These include, for example, the complaint that many software providers charge for technical assistance. Providing technical assistance has a cost. A software provider can allocate that cost to those who use the service by charging for the service, or can allocate the cost to everyone who uses the software, regardless of whether or not they use the technical assistance, by raising the cost of the software for everyone. Insisting that a certain number of hours of technical assistance be provided with all products effectively requires software providers to allocate the cost of such assistance across all of the users of their products, rather than to allocate the cost exclusively to those who use the service. Again, no rationale is offered for the proposition that it is appropriate for legislation, rather than the market, to determine whether technical assistance is required, and who should pay for it. Insisting that a uniform law require that technical assistance be provided is likely to impose the greatest burden on the small developer who is least able to bear that burden. Again, such an arrangement favors large providers who have many products, and places what may be an insurmountable burden on small developers who have only one or a few products, which may or may not be able to support the minimal amount of required "free" technical assistance. Mr. Kaner claims to represent small businesses, but his proposals, if adopted, would be enormously burdensome to the many small businesses which have developed and are attempting to commercialize software.

Recent decisions have consistently upheld shrinkwrap licenses and similar contractual arrangements.^[22] If innovation and growth of an industry are an indication, the licensing technique has served the industry, and society, well. Competition in the software industry is keen. We have, as potential users, a choice of thousands of "off-the-shelf" software applications. License provisions have not, to date, prevented the creation and distribution of competitive applications.

Mr. Kaner contends otherwise, citing MAI v. Peak.^[23] Reference to the MAI case is somewhat misleading, as MAI v. Peak did not involve a "shrinkwrap" license. In that case, MAI offered to its customers, the installation and maintenance of the MAI computer system, as a package, and priced the package accordingly. A third party service organization, Peak, offered to service the MAI system, and MAI objected, on the grounds, inter alia, of breach of contract. MAI proved to the court that it would not have provided the system unless it also obtained a contract to service the system because it made its money from servicing, not installation (very much like Gillette offers razors for free or a very small price and makes its money off of the razor blades). On that basis, the court upheld the provisions of the license agreement which precluded servicing by third parties. It is worth noting that even though the court upheld MAI's legal position by upholding the enforceability of its license agreement, MAI's customers appear, ultimately, to have found the arrangement unacceptable. MAI no longer exists as a separate company. Apparently, the contractual obligations proved a poor defense against the strong competitive forces in the industry. Despite MAI's legal victory, it is no longer in business, and the business of servicing computer systems remains highly competitive. The moral of the story (which is not lost on licensors and their lawyers): Customers unhappy with a licensor's arrangements are likely to seek, and to find, substitute arrangements more to their liking.

To the extent that abuses exist in the software industry, remedies exist under current law. For example, to the extent that software publishers fail to provide what they promise, the remedy lies in a breach of contract claim, or possibly, an action for fraud in the inducement, in which case the contract is unenforceable. If advertising is false, deceptive and misleading, remedies are available under Section 42 of the Lanham Act and state statutes prohibiting unfair and deceptive trade practices.

Article 2B is, like the original Uniform Commercial Code, an attempt to codify existing law and practice, resolve conflicts of law where they have developed, and facilitate trade and commerce. It is far from perfect.

The legislative process invites affected groups to attempt to influence legislation to their benefit. It provides an opportunity for individuals and groups (e.g. Mr. Kaner and like-minded others) to influence a market in ways which the market system, as an economic system, does not permit. Where the market system appears not to be operating satisfactorily, government intervention may be appropriate. Where the market system is working, or other avenues for appropriate government intervention exist when it is not working, permitting proposed legislation to become a basis for negotiating more favorable terms for certain interests than they can now obtain in the market place is, at least in a market system, ill-advised. Such negotiation through legislation is particularly undesirable when these interests make demands without regard to their second-level effects, which frequently take the form of unintended (and therefore unconsidered) consequences of government intervention.

For software developers, publishers and users ("consumers"), the question is not whether Article 2B is ideal, but whether, on balance, those who provide and utilize software, and society as a whole, would be better off with the proposed law in place than without it. That is the fundamental question which the Drafting Committee will ultimately have to decide and the appropriate question for determining whether proposed Article 2B deserves support.

The software industry has found that the licensing model works well. It offers maximum flexibility in an industry in which change is constant and rapid. To the extent that a uniform law confirms the enforceability of contractual arrangements, it is likely to be welcomed. To the extent it reduces flexibility, makes enforcement of contracts less predictable, or potentially narrows efficient and effective distribution channels, it presents a potential for disruption of the market and diminution of opportunity, and is therefore less likely to be welcomed by the industry.

FOOTNOTES

1Copyright 1998, Micalyn S. Harris; All Rights Reserved. Cite as: Micalyn S. Harris. "Is Article 2B Really Anti-Competitive?" 3 Cyberspace Lawyer No. 8, November, 1998. Text may vary slightly from printed version.

2"In My Opinion: Restricting Competition in the Software Industry - The Impact of Pending Revisions to the UCC", 3 Cyberspace Lawyer No. 3, p. 11.

3Article 2B of the Uniform Commercial Code may be downloaded from http://www.law.uh.edu/ucc2b/.

4According to the United States Department of Commerce, the world packaged software market was $109.3 billion in 1996, of which $50.4 billion was in the U.S. The world packaged software market is expected to exceed $125 billion in 1997. U.S. Industrial Trade Outlook '98, United States Department of Commerce, 1998. Note that these figures reflect only packaged software. The scope of Article 2B is not limited to packaged software, and therefore, Article 2B would impact an even larger market.

5The author is General Counsel of Winpro, Inc., a relatively small software consulting, design and development company which offers both "shrinkwrap" software and custom consulting, system design and software implementations.

6See 17 U.S.C. Sec. 102

7See 17 U.S.C. Sec. 107

8See e.g., "Free Speech, Fair Use and the Joys of Sects: The Evolving Scope of Copyright Protection on the Internet", The DataLaw Report, November, 1996.

9Although personal computers, and some software, were available from Tandy and Apple, the market began to broaden and grow meaningfully only in 1982, with the introduction of the IBM PC, giving us less than twenty years of case law regarding these issues.

1017 U.S.C. Sec. 107

11See e.g. Borland's license for its Turbo-Pascal and C ++ software.

12UCC 2B-208, draft of August, 1998.

13See UCC 2B-208, draft of August, 1998.

14See Lewis Galoob Toys, Inc. v. Nintendo of Am., Inc., 964 F.2d 965 (9^th Cir. 1992).

15See discussion in ProCD, Inc. v. Zeidenberg, 908 F. Supp. (WD Wis. 1996), and Nimmer, Raymond T., Information Law, Para. 2.08[3][b].

16Lewis Galoob Toys, Inc. v. Nintendo of Am., Inc., supra, see also, Sega v. Enters. Ltd. v. Accolade, Inc., 977 F2d 1510 (9th Cir. 1992) and Atari Games Corp. v. Nintendo of Am., Inc., 975 F2d 832 (Fed. Cir. 1992).

17See e.g., Andrew Schulman, David Maxey and Matt Pietrek, "Undocumented Windows: A Guide to Reserved Microsoft Windows API Function" (1992).

18As a practical matter, a customer is permitted to return the unopened diskette for any reason, not just unhappiness with the terms of the license, and the provisions of the August, 1998 draft of 2B-208 reflect this reality.

19"Clickwrap" licenses may be included in packaged software, and are displayed at the beginning of the installation process. Software which is distributed electronically, that is, never packaged, also generally includes a "clickwrap" license to which a customer must agree before being permitted to download the software.

20See 2B-208 (b), draft dated August, 1998. Mr. Kaner equates "shrinkwrap" with "clickwrap" licenses and finds "clickwrap" licenses as unacceptable as "shrinkwrap" licenses. If, however, review of license terms prior to shipment is the objective, "clickwrap" arrangements should meet Mr. Kaner's criteria, as review prior to ordering is easily arranged. He does not discuss why is objects to "clickwrap" licenses. If, however, preshipment review of license terms were required, it would support electronic distributors to the detriment of catalog distributors. As a matter of public policy, in the absence of discussion and evidence of a need for legislation to support one distribution method rather than another, we believe it is preferable for the marketplace, rather than state legislatures, to make that determination.

21Kaner, supra, at 13.

22See e.g. Pro CD v. Zeidenberg, 86 F. 3^rd 1447 (7^th Cir.1996) and Brower v. Gateway 2000, Inc., 1998 WL 481066 N.Y.AD1Dept. (Aug. 13, 1998).

23MAI Systems Corp. v. Peak Computer, Inc., 991 Fed2d 511 (9^th Cir.1993).