Lapas attēli

would result.

5.7.2 Computer Programs and Literary Works

Seven categories of works are now granted protection under Section 102 of the 1976 General Revision of Copyright Law. While the definition of "literary works" given in Section 101 of the new Act is broad enough to include computer programs, it is not necessary that computer programs be defined for purposes of the statute as literary works. An alternative is a new category of copyrightable work to be enumerated in Section 102, namely "computer programs.

One reason for consideration of this question is that computer programs are used in different ways than prose or poetry. The limitations on exclusive rights granted to users of literary works, for example, as specified in Section 110 of the 1976 General Revision, may or may not be appropriate for computer programs. In particular, the applicability of the limitations of Section 110 to computer programs used for computer-assisted instructional purposes is worthy of examination.

Similarly, as the uses to which computer programs are put or the manner in which they are used differ from more standard literary works, additional modifications of the copyright statute may be appropriate to specify the assignment of property rights with respect to each type of work. Categorization of computer programs separately from literary works might assist the process of specifying these differences.

5.7.3 Originality of Computer Programs

While no specific research study can be identified yielding definitive results that computer programs can be "original", as the meaning of that term is understood in copyright law, experience and knowledge of the field make possible an unequivocal affirmative response.

Many books have been written on the subject of how to write programs and how to write better programs. If originality were not possible, it would have been difficult if not impossible for Gerald M. Weinberg to have written the book The Psychology of Computer Programming 69 including sections on "Programming as Human Performance" and Programming as an Individual Activity." Similarly, it would have been far less likely for Dennie Van Tassel to have written on "Program Style" in his book on Program Style, Design, Efficiency, Debugging, and Testing?" or for Frederick P. Brooks, Jr. to have written of "the joys of the craft" or of "craftsmanship" in his book on The Mythical Man-Month, Essays on Software Engineering.


Of course, the more complex a program's function, the greater the variety of unique ways of expressing the steps in the performance. On the other hand, it is questionable whether a program carrying out an elementary and well-defined function such as the calculation of the roots of a second-order polynomial could be considered "original." It may be

within the discretionary power of the Register of Copyrights to deny copyright to such a program on that basis. It is likely, however, that the copyrighting process will be self-regulating. Only programs having an intrinsic originality are likely to be submitted for registration.

5.7.4 Protection of Object Code as a Computer Program

The object code is the conversion into symbols usable directly by the computer of the source program written by the programmer. The basic question with respect to object code is whether it should be able to be copyrighted independently of the source code. If it were independently copyrightable as a computer program, a programmer could submit the object code to the Copyright Office for registration and never disclose the source code at all.

The point has been raised that, very likely, the sequence of ones and zeros in hard-copy form constituting the object code is, in the abstract, already copyrightable as a literary work under present law. Analogously, the sequence of numbers in a data base are clearly copyrightable and similarly, original sequences of nonsense syllables are acceptable for registration since no value judgment need be made as to literary merit.

However, the concept of a "computer program" implies a sequence of instructions involving a solution to a quantifiable problem. The granting of the protection of copyright implies the right to prevent infringements and imposes responsibilities on the Government. Yet the object code (except for a program of very short length) is unreadable as a computer program by a person. It would be exceedingly difficult for the Copyright Office to assure that the object code was "original" for registration purposes and similarly difficult for the facts to be determined in an infringement action.

The registration of the sequence of ones and zeros constituting the object code could be used, certainly, to prevent unauthorized copying and use of exactly that sequence. However, many infringements of the underlying program could occur without the use of the exact sequence. For example, it would be extremely easy to shift the specific sequence while still plagiarizing the program through the insertion of a single instruction not changing the logic of the sequence, or to change the encoded addresses of operands, or to use different encodings for the machine commands. A copyright registrant might find that object code registration actually provided, as a practical matter, very little real protection.

In addition, copyright registration of object code as a computer program discloses almost nothing in return for the protection of law. Information transfer about the program is deliberately minimized, not maximized. Thus, this study finds that the independent copyrightability of object code as a computer program is not in accord with the basic principles on which its recommendations are based.

On the other hand, the above should not be understood as implying the finding that object code is not protectable at all. The copyrightability of programs in source language would have very little value if the object code could be produced or copied with impunity. It is concluded, therefore, that the conversion of a source program into object code, which implies no addition to the logic of the program and therefore no value added, constitutes the making of a copy.

Thus, object code should be protected by virtue of the copyright in the source program. It may be noted that in the process of producing object code from a source program, the usual procedure is to combine certain necessary operating parameters into the object code. These parameters often select the specific peripheral units that will be used with the program when the program is run and also select the location of the program in the computer storage units. In the view of this study, these additions to the object code constitute almost nothing that could be classed as original works of authorship. Thus, the generation of object code, even with the addition of these housekeeping functions, cannot be classed as the preparation of a derivative work.

5.7.5 Translation to a New Source Language

The translation of a source program from one source language to another source language should be considered the preparation of a derivative work. The translation makes possible the understanding of the program by an additional group of persons and provides for wider dissemination and use.

[blocks in formation]

It is clear from the concept of copyright and from Section 102(b) of the 1976 General Revision that only the "expression" of a program can be protected. As stated in Section 102(b):

"In no case does copyright protection for an original work of
authorship extend to any idea, procedure, process, system,
method of operation, concept, principle, or discovery, re-
gardless of the form in which it is described, explained,
illustrated, or embodied in such work.

The question may be asked, whether protecting the expression only, rather than the concept is valuable. An answer is that copyright protection hopes to prevent a major type of market failure with regard to computer programs, but does not claim to protect against all types of market failure. Therefore, copyright is valuable, but not valuable for every purpose.

It is important to note that unauthorized copying of computer programs, even without any further use or dissemination of the concepts of the program, is a major type of market failure. The reason this is true is that examination of the program code to determine any unique concepts

contained therein requires the expenditure of significant resources, while copying by itself requires only a bare minimum of resources. A copier who is assured that the program in question performs the functions he desires in an error-free manner has obtained something of considerable value, at minimum expense. The added effort of understanding any unique procedures contained in the program is not likely to yield a corresponding advantage for a pragmatic user.

The disclosure of unique concepts, certainly, will assist competitors in the development of competing programs, but whether a particular unique or innovative design concept is protectable would depend on how a statute (such as the patent law) protecting such concepts might be written or might be interpreted. This report is not the proper vehicle for a detailed discussion of this matter; but it can be pointed out that very few programs contain (or need to contain) new concepts as unique as the simplex method for the solution of linear programming problems or the fast fourier transform algorithm, both outstanding advances in computational procedures. For the most part, what is required of programs is that they carry out their intended functions with precision and in an error-free manner. Performance is improved if in addition, programs minimize execution time and use of storage space to the extent practicable. For most applications, unique concepts are not required, and for these programs, copyright protection should be sufficient. Clearly, there appears to be room for further study on the possible protection of unique and innovative programming concepts.

5.7.7 Copyright and Software Industry Strength

One argument against copyrightability of computer programs is that the industry is burgeoning and therefore copyright is unnecessary. It must be noted, however, that copyright does not specifically protect an industry, but rather a particular work in the marketplace. The protection is particularly important for the smaller entrepreneur who does not have the resources to engage in the kind of retaliatory measures suggested by Hurt and Schuchman or to protect himself against the predatory practice proposed by Breyer and described in Section 4.2 above. Copyrightability promotes competition and innovations by the small competitor. These aspects of the marketplace are important criteria for public policy towards an industry, as are growth and size of the industry. 5.7.8 Duration of Copyright Protection

It seems reasonable to propose that the author of a computer program should not be treated any differently than the author of any other type of copyrightable work. Therefore, the duration of copyright in computer programs should be the same as the duration of copyright in other works.

A reason that has been given for proposing a shorter duration of copyright is that with changing technology, computer programs would become valueless after several years. However, if the proposal of this report is adopted, that an original computer program copyright should be

obtainable only in the source program, and not in the object code, then a separation of the programmer's expression from the hardware technology is promoted. Furthermore, even if popular source languages are altered or improved, the copyright proprietor retains the right to prepare derivative works, permitting him to update the program as required.

5.7.9 User Rights in Computer-Readable Works

A computer program, and a computerized data base as well, are intended for use in conjunction with a computer. That is, a computer-readable work is used by entering it into a computer system and manipulating it through the logic of a computer. It seems reasonable to propose that the copyright proprietor should retain the exclusive right to the use of a computer-readable work in a computer.

However, this study proposes a limitation on the exclusive right of use, in order to reduce transaction costs in connection with the transfer of ownership of copies of computer-readable works. This limitation is intended to operate through improved salability of computer programs and computerized data bases, considered immediately below.



Several kinds of copyrighted works are offered for sale at retail. Books, maps, and sound recordings are typical of this class. The advantage of sale over lease or rental is that transaction costs are minimized. No agreement, except to pay the retail price, need be made. The buyer obtains ownership over the copy or phonorecord he has purchased, including the right to resell that copy, except for certain rights retained by the copyright owner. The retained rights include the rights to make and sell copies (with exemption for fair use, compulsory licenses, etc.), the right to prepare derivative works, and the rights to perform and display the work publicly.

If the rights to computer-readable works could be defined in such a way as to promote the sale rather than lease of such works, transaction costs might be similarly minimized. This would be, certainly, in the public interest.

[blocks in formation]

One of the problems in the sale of computer-readable works is the right of the buyer to copy for his own use. Here, "buyer" means the purchaser of a copy where ownership of the copy is transferred. For works published in paper, "use" simple means "reading" and no copying is required. For sound recordings, "use" means "playing" the recording on a playback mechanism, but again, no copying is required. For computerreadable works, copying into the computer is required in order to use, and in addition, archival copies are made in normal practice in case a copy in use is destroyed inadvertently.

« iepriekšējāTurpināt »