XX XXXXXX XXXXX XXXXX XX XX XX XX XX XX XX XX XXXX XX XXX XXXXXXX XX XX XX XX X XX XX XX XXXXXX XXXXXX XXXXX XX XX XXXXXX XXXXX XX XX XXXXXX XXXXXX XXXXXX XX XX XX XX XX XX XX XXXXX XXXX XX XXXX XXX XX XX XX XX XX XX XXXXX XX XX XXXXXX XXXXXX __________________________________ Winter 1992-93, Volume 1, Number 1 __________________________________ By George, Donaldson & Ford Attorneys at Law A Registered Limited Liability Partnership 1000 Norwood Tower 114 West Seventh Street Austin, Texas 78701 (512) 495-1400 (512) 499-0094 (FAX) ___________________________________ Copyright (c) 1993 George, Donaldson & Ford, L.L.P. (Permission is granted freely to redistribute this newsletter in its entirety electronically.) ___________________________________ David H. Donaldson, Jr., Editor Peter D. Kennedy, Associate Editor ___________________________________ IN THIS ISSUE: 1. ARE ELECTRONIC BULLETIN BOARD SYSTEM OPERATORS LIABLE FOR THEIR USER'S LIBELOUS STATEMENTS? 2. ELECTRONIC PUBLISHING, BULLETIN BOARD E-MAIL, AND THE STEVE JACKSON GAMES CASE 3. WHEN IS A COMPUTER PROGRAM A "COPY"? 4. CAN YOU INFRINGE A COPYRIGHT WHILE ANALYZING A COMPUTER PROGRAM? 5. DOES ANTITRUST LAW HAVE A PLACE IN THE COMPUTER INDUSTRY? 6. WHAT DO YOU MEAN WE DON'T OWN THE SOFTWARE? ========================================================================= 1. ARE ELECTRONIC BULLETIN BOARD SYSTEM OPERATORS LIABLE FOR THEIR USER'S LIBELOUS STATEMENTS? Electronic bulletin boards are now very popular "places" for people to "meet" and discuss issue of the day -- from politics to fantasy game- playing to bird watching. The unique freedom bulletin boards offer is both a great advantage and a potential source of new legal questions and headaches. Many bulletin boards are free of charge, or charge only modest fees, and the cost of a personal computer and modem has dropped dramatically. Unlike other means of mass communication, electronic bulletin boards present no significant barriers to entry; almost anyone can become a fully active member in wide-ranging discussions and debates involving hundreds or even thousands of participants and observers. Users are free to post whatever messages suit their fancy, limited only by the possibility of editing or deletion by the system operator. The system operators of most bulletin boards commonly do not edit or control the messages and files posted by users. For the large communication services such as CompuServe and Prodigy, meaningful editorial control or review of messages and uploads is impossible. The system operators of smaller boards generally prefer to allow the discussions on their boards to range freely, while simply defining the areas and topics for discussion, and perhaps setting the tone or direction of discourse. The technical work of maintaining a board is demanding enough, without the added chore of reviewing the content of each posting and file. Many sysops are rightly satisfied to provide a forum for discussion -- a democratic, uninhibited common ground in cyberspace for the "robust and wide-open" debate of public issues the First Amendment is meant to guarantee. Where technology leads, law is sure to follow. A host of legal questions have arisen around electronic bulletin boards, from the legal protection for private electronic mail carried on boards, to the right of BBSs to discount telephone rates. Many legal questions involve the sysop's possible liability, both civil and criminal, from potential criminal liability for knowingly providing a means for the distribution of electronic "contraband" (such as purloined passwords, or stolen documents, codes, and credit card numbers) to possible civil penalties for copyright infringement. A significant legal question is how the age-old civil lawsuit of defamation of character should apply (if at all), to the brand-new medium of the electronic bulletin board -- more specifically, when is the system operator of a bulletin board liable for defamatory statements posted by a user? Despite a spate of articles about the legal protection, constitutional and otherwise, enjoyed by sysops and BBS users, very few actual court decisions have been issued defining the scope of protection and potential liability. The most significant decision to reach the law reports so far concerning sysop liability for defamation is CUBBY, INC. v. COMPUSERVE, INC., (reported in 776 F. Supp. 135 (S.D.N.Y. 1991)). CompuServe Inc., which runs one of the larger on-line information systems, included in its Journalism Forum a daily newsletter called "Rumorville USA." A company called DFA wrote and published the newsletter, and provided it to a company known as CCI. CCI had contracted with CompuServe to manage, review, and edit the contents of the Journalism Forum, and uploaded "Rumorville USA" to the service, without prior approval or review by CompuServe. CompuServe had no contract with DFA, the newsletter's author, but only with CCI. Cubby, Inc., a company that was developing a journalism news database hoping to compete with "Rumorville USA," was unhappy with a story in a "Rumorville USA" issue that mentioned Cubby. Cubby sued DFA and CompuServe, alleging that the article libelled Cubby and its owner. Judge Leisure of the federal district court located in New York City held that even if the "Rumorville USA" article defamed Cubby (which was in dispute) *CompuServe* could not be held responsible. The judge saw CompuServe as an "electronic, for-profit library," that had no opportunity to review the content of its holdings before they are published. "Secondary publishers" such as libraries and bookstores are not responsible for the content of the books they offer unless they know, or have reason to know, of the books' objectionable nature. The judge noted that the sheer volume of the material posted on CompuServe kept it from ever knowing or having reason to know that a certain posting was defamatory. Because the technology and volume made it unreasonable and unfeasible for CompuServe to monitor every statement made on its service, Judge Leisure dismissed the suit against CompuServe. While sysops breathed a sigh of relief after CUBBY v. COMPUSERVE was decided, the case will not be the last word. For one, while other courts will pay it some mind, the case is controlling precedent for only one federal trial court in New York City. New York's status as an important center of both libel *and* computer law does increase the persuasive authority of CUBBY v. COMPUSERVE, however. Second, the specific facts of the case keep CUBBY v. COMPUSERVE from being a sweeping precedent. Because CompuServe is so big, it is reasonable not to expect it to review the contents of every posting, but what about smaller boards? If the system operator of a board is also an active user, and could theoretically review all the latest postings, is he or she protected by the decision in CUBBY v. COMPUSERVE? What if CompuServe had been given notice of the defamatory nature of the article, but refused to remove it from the system? Judge Leisure's logic would indicate that if there was proof that CompuServe *knew* that "Rumorville USA" had defamed Cubby, it would not escape liability on that ground alone; it would have to defend the defamation action on more traditional grounds -- that the statement was a constitutionally protected statement of opinion on an issue of public concern, or a privileged comment about a public figure, for example, or that the statement, although harmful, was substantially true. (And, there is currently little reason to think that law of libel will apply any differently to the person *who* actually makes the defamatory statement on an electronic bulletin board, than it does to a person making such a statement. Judge Leisure has not published any opinion discussing the outcome of Cubby's suit against the *authors* of "Rumorville USA.") Paradoxically, Judge Leisure's decision in CUBBY v. COMPUSERVE may encourage sysops *not* to review the contents of their boards, in order to claim ignorance of defamatory postings. Outside the courtroom, writers have been trying to develop legal theories that give system operators satisfactory protection. These range from the "secondary publisher" analogy adopted by CUBBY v. COMPUSERVE, (see L. Becker, Jr., "The Liability of Computer Bulletin Board Operators for Defamation Posted by Others", 22 Conn. L. Rev. 203 (1989)), to suggesting that electronic bulletin boards might be treated like "common carriers" such as telephone companies, making them immune from any liability for statements passing through their wires (see R. Charles, "Computer Bulletin Boards and Defamation: Who Should be Liable? Under What Standard?", 2 J. Law & Tech. 121 (1987) and E. Jensen, "An Electronic Soapbox: Computer Bulletin Boards and the First Amendment," 39 Fed. Comm. Law J. 217 (1987)). An upcoming article even argues that the unique nature of electronic bulletin boards should *protect* the sysops, not present additional legal risks: because bulletin boards, unlike other mass media, offer persons who think they have been defamed by a message on the board a free, immediate, and practically unlimited right to reply to the message, the law should recognize a privilege for the sysop from suit for defamation carried on his or her board who allows access for replies. This would encourage rather than stifle free discussion, the purpose of the First Amendment (see E. Cavazos, Note, "Computer Bulletin Board Systems and the Right of Reply: Redefining Defamation Liability for a New Technology," Volume 12, The Review of Litigation (forthcoming)). If the experience of the more traditional mass media is any indication, we can expect more defamation lawsuits brought against the operators and owners of bulletin board systems, until the contours of the law are better settled by concrete decisions or legislation. _________________________________________________________________________ 2. ELECTRONIC PUBLISHING, BULLETIN BOARD E-MAIL, AND THE STEVE JACKSON GAMES CASE Currently moving through federal court is a lawsuit that may help define the legal protection for electronic publishers and electronic mail, and set limits on government law enforcement officials searching for evidence of illegal computer hacking. On May 1, 1990, the United States Secret Service (charged with investigating computer crime as well as protecting the President) searched the offices of Steve Jackson Games, a fantasy role-playing game publisher in Austin, Texas. The Secret Service confiscated and kept for months three of the company's computers, including one computer that ran the company's electronic bulletin board, and another that contained drafts of an upcoming book. The Service claimed it was investigating suspected illegal activity of a Steve Jackson Games employee. In May, 1991, Steve Jackson, his company, and three active users of the company's bulletin board sued the Secret Service and the responsible officials to recover the damages caused by the raid, which included the temporary loss of the bulletin board, as well as the loss of the book. Prominent issues in the lawsuit include whether the Secret Service had probable cause to search the offices in the first place, and whether the government violated the Electronic Communications Privacy Act (which prohibits the illegal interception, disclosure or use of electronic communications) by seizing and reviewing the BBS computer. In recent filings, the government asked the court to dismiss the case, arguing, among other things, that the bulletin board's users had no "reasonable expectation of privacy" in the contents of their e-mail, that Steve Jackson's book-form, role-playing games and magazines are not worthy of protection by the Privacy Protection Act, and that the Electronic Communications Privacy Act does not prohibit the government from seizing and reading the entire contents of a BBS, including the e-mail, so long as there is a search warrant which authorizes the seizure of computers. Steve Jackson and the three users of his board (represented by George, Donaldson & Ford) have asked the judge to rule that even as the Secret Service tells the story, the law was violated. The case was tried before Judge Sam Sparks in an Austin, Texas federal court the week of January 26, 1993, and awaits Judge Spark's ruling. _________________________________________________________________________ 3. WHEN IS A COMPUTER PROGRAM A "COPY"? When the computer whizzes of today were growing up, one of the most popular Saturday morning cartoon programs was "H.R. Pufnstuf." This creation of Sid and Marty Krofft - a life-action adventure featuring human caricatures and fanciful creatures - lead to one of the first modern cases to grapple with an issue that now plagues the computer software industry: What is the dividing line between an idea, concept, or process, which is not protected by copyright law, and the expression of an idea, which is protected? When McDonald's introduced its "McDonaldland" with "Hamburglar," "Fry Guys," talking fries with human faces, and other characters, Sid and Marty Krofft sued, alleging that although the McDonaldland creatures were not exact imitations of the characters in "H.R. Pufnstuf," they were substantially similar and therefore their use constituted a copyright infringement. (SID & MARTY KROFFT TELEVISION v. McDONALD'S CORP., 562 F.2d 1157 (9th Cir. 1977)). A jury, after reviewing both creations, agreed with Sid and Marty, and the courts ultimately affirmed the jury's verdict, holding that although "Hamburglar" and his buddies were not exact duplicates, they "captured the 'total concept and feel'" of "H.R. Pufnstuf," which justified a finding of substantial similarity and, in turn, copyright infringement. McDonald's Corporation was forced to seek a license from the Kroffts. The same scenario, although now much more complex, is being played out in the computer industry. Over the last seven years federal courts, which consider cases involving the copyright infringement, have been grappling with what constitutes copyright infringement in the software industry. These cases arise because as new types of programs are created or new ways of doing or presenting the same type of function are adopted, persons who believe that they can produce a product that offers the same functionality with improvements want to get into the market as well. Not surprisingly, because consumers are reluctant to re-learn programs, new entrants into the market try to make their product familiar to those who have become accustomed to the industry standard. New entrants may use some of the key strokes, screen displays and other characteristics of the industry standard, along with the innovations and advantages that the new entrant offers. Normally the new entrant's software has different source code and object code from the industry standard. So there is no *literal* copying of the source code or object code of the competitor. However, when a new entrant incorporates some of the screen displays, the methods in which the program operates, or the key strokes of the competitor, some courts have held that this *non-literal* copying can be copyright infringement as well. The court attempting to sort out the legal differences and similarities between "H.R. Pufnstuf" and McDonaldland had it easy compared with today's courts that are struggling with a way to express in legal terms the standard that will apply in deciding whether a software program is a "copy" of another program. Three schools of thought are currently battling for supremacy in the courts throughout the country. The first two approaches seem to be losing ground, to, naturally, the one most complex and difficult to apply. The first concept, prevalent primarily in the southwest in the United States Court of Appeals for the Fifth Circuit (which includes Texas, Louisiana, Mississippi) essentially says that absent literal copying of the code or the screen displays, similarity in the structures of the program, the sequence of key strokes, and other methods by which the program operates are not copyright infringement. (PLAINS COTTON CO-OP v. GOODPASTURE COMPUTER SERVICE, INC., 807 F.2d 1256 (5th Cir.), cert. denied, 484 U.S. 821 (1987)). This approach was announced early in the development of this area of the law. While it affected the development of the law in this area, it does not appear to have gathered very many adherents. The second approach, used by the United States Court of Appeals for the Third Circuit (Pennsylvania, Delaware, New Jersey, Virgin Islands) provides expansive copyright protection to computer software. In WHELAN ASSOCIATES, INC. v. JASLOW DENTAL LABORATORY, INC., 797 F.2d 1222 (3rd Cir. 1986), the Third Circuit analyzed whether a dental program designed for the IBM Series I computer written in event-driven language ("EDL") was infringed by the defendants' "Dentlab" program designed for the IBM/PC computer and written in Basic. Obviously, the two programs were not literally the same because they had different languages in which they were written, so the question was whether the *non-literal* elements of the program -- the structure of the program, the sequence of key strokes used to operate the program and the organization of the program -- were substantially similar. The court's key conclusion was that copyright protection of computer programs extends beyond the programs' literal code to their structure, sequence, and organization. The court held that a computer program is a literary work and, like a play or a book, it can be copied by copying its plot or its plot devices even though the copy may not be written in exactly the same words or use exactly the same names. In responding to the claim that the copyright statute does not protect ideas, concepts or processes, the court said that in a computer program the purpose to be accomplished by the program is the "idea" and that everything else that is not necessary to that purpose or function is part of the expression of the idea. Where there are various means of achieving the desired purpose, then the particular means chosen is not *necessary* to the purpose and therefore is expression and not idea. Accordingly, the court compared how the two computer programs accomplished the purpose of the program. It found a virtual identity of most of the file structures used in each program, the screen outputs of both programs, and an almost identical performance of five particularly important sub-routines in the programs. The court affirmed the district court's finding that there was substantial similarity between the programs and therefore a copyright infringement. The WHELAN approach to determining what is protected and copyrightable in evaluating substantial similarity has been under attack by the United States Court of Appeals for the Second Circuit (New York), which serves the important media center of New York. In COMPUTER ASSOCIATES INTERNATIONAL, INC. v. ALTAI, INC., 61 U.S.L.W. 2434, 1992 WL 372273 (2d Cir. 1992) (opinion on rehearing), the court adopted a different approach to analyzing substantial similarity: a three part test involving abstraction, filtration and comparison. The ALTAI court began by explaining how it understood a computer program was created: first, the programmer must identify the program's ultimate function or purpose; second, the programmer breaks down the ultimate function into subtasks; third, the subroutines are organized into flow charts outlining the interaction (the structure of the program); fourth, to assure the subroutines' proper interaction, parameter lists are prepared; and finally, once the modules have been identified, designed and interrelationships laid out conceptually, the code must be written to reflect the design. Source code is written in English language statements and then compiled into object code. The court noted that the actual "text" of a computer program -- the source and object code -- are firmly protected by copyright law. The court held that computer programs are "literary works" under the Copyright Act, and therefore some non-literal aspects are protected as well, but only so far as they incorporate the programmer's *expression* of ideas as opposed to the *ideas* themselves. The Second Circuit said the first step in deciding whether a program which does not literally copy source or object code is "substantially similar" to a copyrighted program is "abstraction." In essence, the court reverse engineers the copyrighted computer program theoretically by dissecting the allegedly copied program's structure and isolating each level of abstraction contained within it. The process begins at the most basic level with the code and ends with the articulation of the program's ultimate function -- for example, a spreadsheet or a database. The court retraces and maps the designer's steps but in the opposite direction taken by the designer. Once the program's abstraction levels are determined, then a process of "filtration" occurs. In the filtration process the structural program is "filtered" through a legal analysis to take out all of the aspects of the program that are not protectable. The "filters" include (1) whether a structural component was an idea or dictated by considerations of efficiency so as to be necessarily incidental to that idea (the "idea" of columns on a spreadsheet would include the code necessary to create the columns); (2) whether the structural component is required by factors external to the program itself (all spreadsheets need columns); or (3) was taken from the public domain and hence, is nonprotectable expression (the use of columns in accounting). To elaborate, if there is only one efficient way to accomplish the idea, then the procedure or method used to accomplish the idea merges into the idea itself and is not protectable. Elements dictated by external factors involve certain standard or stock devices that must be used to make the program work. Examples of such extrinsic considerations include (1) the computer's mechanical specifications; (2) compatibility requirements of other programs; (3) the computer manufacturer's design standards; (4) demands of the industry being serviced; and (5) widely accepted programming practices within the computer industry. Elements taken from the public domain would include material that is available generally such as shareware, standard programming techniques, or code generators. The final step is "comparison." Once the court filters the program that was allegedly copied to take out all of the non-protected elements, it compares the remaining nugget of protected expression with the allegedly infringing program to see if there is any substantial similarity between the two. After the ALTAI court completed this process, although there were some similarities between the programs, they were so small or insignificant, or so simple and obvious that their use could not be considered copyright infringement. The ALTAI court recognized that its opinion would make it substantially more difficult for the owners of computer software to assert copyright infringement actions against infringers. The court seemed to confess that it had not clarified matters much: "to be frank, the exact contours of copyright for non-literal program structure are not completely clear" and anticipated that future cases would define the parameters more clearly. ALTAI's approach differed from WHELAN because in ALTAI the court examined *each* level of abstraction in the program as opposed to the overall program in its entirety. The WHELAN court's approach was considered to be too overbroad because it found that the program had only one "idea" and everything else was considered "expression." The ALTAI court considered its more sophisticated abstraction, filtration and comparison tests superior. Subsequent courts appear to have sided with ALTAI and against WHELAN. In ATARI GAMES CORP. v. NINTENDO OF AMERICA, INC., 975 F.2d 632 (Fed. Cir. 1992), the United States Court of Appeals for the Federal Circuit, the court that considers patent and related lawsuits, adopted the ALTAI approach to analyzing substantial similarity. Similarly, the United States Court of Appeals for the Ninth Circuit, which includes California, has also endorsed the dissection of copyrighted work to determine the level of protection that will be afforded. BROWN BAG SOFTWARE v. SYMANTEC, 960 F.2d 1465 (9th Cir. 1992). Federal district judges, on the trial level, have also adopted the ALTAI approach to determining substantial similarity. As the Second Circuit admitted, however, the scope of copyright protection for non-literal elements of a computer program remains unsettled. If an exact copy is made of another program then copyright infringement will occur. If a program is a completely independent creation there is no copyright infringement because no copying has occurred. Between those two extremes the law remains muddy and will depend largely on the facts of each case and how the facts are presented to the court.