Protection for Computer Objects

Protection for Computer Objects

The previous sections have described three forms of protection: the copyright, patent, and trade secret laws. Each of these provides a different form of protection to sensitive things. In this section we consider different kinds of computer objects and describe which forms of protection are most appropriate for each kind. Table 11-1 shows how these three forms of protection compare in several significant ways.

Computer artifacts are new and constantly changing, and they are not yet fully appreciated by the legal system based on centuries of precedent. Perhaps in a few years the issue of what protection is most appropriate for a given computer object will be more clear-cut. Possibly a new form of protection or a new use of an old form will apply specifically to computer objects. For example, the European Union has already enacted model legislation for copyright protection of computer software. However, one of its goals was to promote software that builds on what others have done. Thus, the E.U. specifically exempted a product's interface specification from copyright and permitted others to derive the interface to allow development of new products that could connect via that interface.

Until the law provides protection that truly fits computer goods, here are some guidelines for using the law to protect computer objects.

Protecting Hardware

Hardware, such as chips, disk drives, or floppy disk media, can be patented. The medium itself can be patented, and someone who invents a new process for manufacturing it can obtain a second patent.

Protecting Firmware

The situation is a little less clear with regard to microcode. Certainly, the physical devices on which microcode is stored can be patented. Also, a special-purpose chip that can do only one specific task (such as a floating-point arithmetic accelerator) can probably be patented. However, the data (instructions, algorithms, microcode, programs) contained in the devices are probably not patentable.

Can they be copyrighted? Are these the expression of an idea in a form that promotes dissemination of the idea? Probably not. And assuming that these devices were copyrighted, what would be the definition of a copy that infringed on the copyright? Worse, would the manufacturer really want to register a copy of the internal algorithm with the Copyright Office? Copyright protection is probably inappropriate for computer firmware.

Trade secret protection seems appropriate for the code embedded in a chip. Given enough time, we can reverse-engineer and infer the code from the behavior of the chip. The behavior of the chip does not reveal what algorithm is used to produce that behavior. The original algorithm may have better (or worse) performance (speed, size, fault tolerance) that would not be obvious from reverse engineering.

For example, Apple Computer is enforcing its right to copyright protection for an operating system embedded in firmware. The courts have affirmed that computer software is an appropriate subject for copyright protection and that protection should be no less valid when the software is in a chip rather than in a conventional program.

Protecting Object Code Software

Object code is usually copied so that it can be distributed for profit. The code is a work of creativity, and most people agree that object code distribution is an acceptable medium of publication. Thus, copyright protection seems appropriate.

A copyright application is usually accompanied by a copy of the object being protected. With a book or piece of music (printed or recorded), it is easy to provide a copy. The Copyright Office has not yet decided what is an appropriate medium in which to accept object code. A binary listing of the object code will be taken, but the Copyright Office does so without acknowledging the listing to be acceptable or sufficient. The Office will accept a source code listing. Some people argue that a source code listing is not equivalent to an object code listing, in the same way that a French translation of a novel is different from its original language version. It is not clear in the courts that registering a source code version provides copyright protection to object code. However, someone should not be able to take the object code of a system, rearrange the order of the individual routines, and say that the result is a new system. Without the original source listings, it would be very difficult to compare two binary files and determine that one was the functional equivalent of the other simply through rearrangement.

Several court cases will be needed to establish acceptable ways of filing object code for copyright protection. Furthermore, these cases will have to develop legal precedents to define the equivalence of two pieces of computer code.

Protecting Source Code Software

Software developers selling to the mass market are reticent to distribute their source code. The code can be treated as a trade secret, although some lawyers also encourage that it be copyrighted. (These two forms of protection are possibly mutually exclusive, although registering a copyright will not hurt.)

Recall that the Copyright Office requires registering at least the first 25 and the last 25 pages of a written document. These pages are filed with the Library of Congress, where they are available for public inspection. This registration is intended to assist the courts in determining which work was registered for copyright protection. However, because they are available for anybody to see, they are not secret, and copyright registration can expose the secrecy of an ingenious algorithm. A copyright protects the right to distribute copies of the expression of an idea, not the idea itself. Therefore, a copyright does not prevent someone from reimplementing an algorithm, expressed through a copyrighted computer program.

As just described, source code may be the most appropriate form in which to register a copyright for a program distributed in object form. It is difficult to register source code with the Copyright Office while still ensuring its secrecy. A long computer program can be rearranged so that the first and last 25 pages do not divulge much of the secret part of a source program. Embedding small errors or identifiable peculiarities in the source (or object) code of a program may be more useful in determining copyright infringement. Again, several court cases must be decided in order to establish procedures for protection of computer programs in either source or object form.

Protecting Documentation

If we think of documentation as a written work of nonfiction (or, perhaps, fiction), copyright protection is effective and appropriate for it. Notice that the documentation is distinct from the program. A program and its documentation must be copyrighted separately. Furthermore, copyright protection of the documentation may win a judgment against someone who illegally copies both a program and its documentation.

In cases where a written law is unclear or is not obviously applicable to a situation, the results of court cases serve to clarify or even extend the words of the law. As more unfair acts involving computer works are perpetrated, lawyers will argue for expanded interpretations of the law. Thus, the meaning and use of the law will continue to evolve through judges' rulings. In a sense, computer technology has advanced much faster than the law has been able to.

Protecting Web Content

Content on the web is media, much the same as a book or photograph, so the most appropriate protection for it is copyright. This copyright would also protect software you write to animate or otherwise affect the display of your web page. And, in theory, if your web page contains malicious code, your copyright covers that, too. As we discussed earlier, a copyrighted work does not have to be exclusively new; it can be a mixture of new work to which you claim copyright and old things to which you do not. You may purchase or use with permission a piece of web art, a widget (such as an applet that shows a spinning globe), or some music. Copyright protects your original works.

Protecting Domain Names and URLs

Domain names, URLs, company names, product names, and commercial symbols are protected by a trademark, which gives exclusive rights of use to the owner of such identifying marks.