Greg Aharonian
Internet Patent News Service
San Francisco, CA 415-981-0441
(latest version at
October 2001

For over thirty years, software copyright has been a succession of court cases and law review articles based on bad law, bad logic, bad mathematics, and/or bad physics (Benson, CONTU, Whelan and Altai being all of these). I have decided to write a critical review arguing that software copyright (and dependents like TRIPS, GPL, Bernstein, Junger) should be abolished in light of 17 USC 102b and its equivalents - for one reason - it is bad law with no logical basis in the mathematics and physics of information processing.

What follows is a list of over 90 Acts, decisions and law review articles I will be critiquing in the review. A small number of cases are from Asia and Europe, which inherited the problems of the US cases.

Some cases deal with patents, for two reasons. One, any decision detracting from software patentability feeds and supports copyrightability. Second, the 1992 Altai case introduced a test for software copyright that is nothing more than some patent procedures without claims but using different words. The two domains have become one, and should so be discussed.

At the end of this paper is a list of cases ruling, and papers arguing, CORRECTLY, that some or ALL of software is not copyrightable.


White-Smith Publishing Co. v. Apollo (Supreme Court - 209 US 1)

Copying of a song by means of a perforated piano music roll did not infringe the copyright in the musical composition because the music could not be visually perceived from the piano roll. Decided before the era of software, but used by courts to deny the equivalence of source code and object/ROMable code (being used as late as 1979 in the Data Cash decision). In 1909, Congress passed a law that reversed this decision (with further corrections in 1976) on practical grounds not based on science.

Dymow v. Bolton (2d. Circ. - 11 F.2d 690)

A pre-cursor of the bogus merger doctrine used to justify some of software copyright. Here the Second Circuit states "Just as a patent affords protection only to the means of reducing an inventive idea to practice, so the copyright law protects the means of expressing an idea; and it is as near the whole truth as generalization can usually reach that, if the same idea can be expressed in a plurality of totally different manners, a plurality of copyrights may result, and no infringement will exist."

I think not. Multiple expressions of functional ideas should be protected through multiple dependent patent claims, not lazily through copyright. The quote was cited in the 1983 Apple v. Franklin case upholding software copyright.

Copyright Revision Bill of 1965

In 1965, Congress starts deliberations and hearings on revisions to copyright, including computing. It doesn't formally enact the final Bill until 1976. The precursors of confusion over software's functionality start here. The Revision Bill grants copyright protection to an original program, whether on punch cards, tapes, printout or the equivalent. That sounds reasonable, but the Bill left unanswered how far that protection should extend to the practical use of the program, a question still unanswered today (so it is hard to fault anyone back then).

Indeed, Benjamin Kaplan in his excellent 1967 book "An unhurried view of copyright" (amongst other things predicting today's digital rights management software) observes both this confusion of expression and idea, as well as the hardware-software equivalence - "Take a program written with knowledge of a copyrighted program and attaining a similar result by different steps or sequences of steps. A suspicion, indeed, that we ought to be thinking not copyright but patent or perhaps a third quiddity, arises as we are told that the programs, or some of them, can be translated, so to speak, into physical parts of the computer's machinery or circuitry.", which can be done trivially today with FPGAs.

In re Jones (CCPA - 373 F.2d 1007)

Ruled non-functional "printed matter" on functional substrate is patentwise nonstatutory. Used against software patenting until superseded by Alappat (1994), and therefore helped encourage software copyright.

In re Miller (CCPA - 418 F.2d 1392)

Ruled non-functional "printed matter" on functional substrate is patentwise nonstatutory. Used against software patenting until superseded by Alappat (1994), and therefore helped encourage software copyright.

Herbert Rosenthal Jewelry v. Kalpakian (9th Circ. - 446 F.2d 738)

Helped establish the "merger" doctrine, a doctrine which plagued many software copyright decisions, in particular, the discredited Whelan decision. Under the merger doctrine, courts will not protect with copyright the expression of an idea that can be expressed in only a few ways, because doing so would confer a monopoly on the idea itself. Somehow the court missed that multiple forms of expression of functional ideas is why dependent claims where invented.

Critics have warned that people should be suspicious of the merger doctrine because its "indeterminacy overwhelms its logic". The outcome of the merger inquiry, they note, necessarily depends on how the court defines "idea". Since the "idea" of any given work can be defined in an infinite number of ways, they argue, the merger doctrine illuminates nothing. (See Wiley, "Copyright at the school of patent", 1991). In the words of Hemnes ("Three common fallacies in the user interface copyright debate", 1990), the merger doctrine is "bogus".

Claiming anti-merger to assert copyright infringement is only for someone too lazy to patent in a dependent claim a range of functional software ideas with a Markush group.

Gottschalk v. Benson (Supreme Court - 409 U.S. 63)

In the words of Chisum, "The reasoning in Benson is monstrously bad." in which the Supreme Court ruled that computer programs not patentable.

Some of the critical commentary:
"... argues that the Benson rule is unsupported by statute or case law authority or by policy considerations. ... Second, the Benson decision, which held that mathematical algorithms could not be patented, was poorly reasoned and stemmed from an antipatent judicial bias that cannot be reconciled with the basic elements of the patent system established by Congress. ... The reasoning in Benson is monstrously bad." - see {CHISUM}. "The Supreme Court's treatment of [software patenting] also lacked clarity. In Parker and Benson, the Court twice reversed lower court decisions holding that a particular piece of software was protectable under the patent laws." see {LAURENSON}. Another good critique of Gottschalk was written in 1973 (see {DUNNER}), which CONTU ignored in 1978 while fearing Benson. In Chisum on Patents, Chisum also writes "Thus, the second and fourth points (including the 'nutshell') are so illogical that they should be dismissed as possible explanations of the holding in [Benson]."

The CAFC's Judge Rich, just one year after Benson, wondered about the confusing logic of Benson: '"Algorithm" has been used in the sense of a "procedure for solving a given type of mathematical problem" and "formula" is used in the sense of a mathematical formula. The Supreme Court in Benson appears to have held that claims drafted in such terms are not patentable - for what reason remaining a mystery. Under the rules of the legal game, we are obliged to follow its lead as best we can.' (In re Christensen, 478 F.2d 1392)

Shortly thereafter, in In re Johnston (502 F.2d 765), Judge Rich further laments:
"I am probably as much - if not more - confused by the wording of the Benson opinion as many others. .... It seems to me important to focus on what the Supreme Court did in Benson, rather than on the specifics of its explanation of why it did it. I have no idea what was in the collective mind of the six-justice Court in approving the statements:
'It is said that the decision precludes a patent for any program servicing a computer. We do not so hold. * * * It is said we freeze process patents to old technologies, leaving no room for the revelations of the new, onrushing technology. Such is not our purpose.'

Judge Rich, commented a few years later, 'In the Benson opinion, the Court states at the outset that "The question is whether the method described and claimed is a 'process' within the meaning of the Patent Act." (Sections 101 and 100(b) were quoted in a footnote to that statement.) But nowhere in the Benson opinion have I ever, after many attempts, been able to find an express answer to that question, ....' (In re Chatfield, 545 F.2d 152) (To be fair, I should note that the lower court decision reversed by the Supreme Court in Benson was written by Judge Rich.)

Benson quotes from the 1966 Presidential Commission on the Patent System that recommended against patent protection for computer programs (a quote Chisum argues is the only way to justify Benson). Benson quotes three paragraphs of the Commission's report. What is little mentioned is that the 1966 Commission was dominated by IBM, which at that time used software to sell its hardware, and didn't want its hardware sales disrupted by software patents (a policy IBM later totally rejected, becoming the biggest software patentee in the world).

Ignore anyone who mentions Benson without mentioning these many problems.

Copyright Act of 1976

First introduction of software copyright into federal law. The basis for doing so was confusing Congressional intent from the prior year (Rep. No. 473, 94th Congress, 1st Session 54 [1975]):
Section 102b is intended, among other things, to make clear that the expression adopted by the programmer is the copyrightable element in a computer program, and that the actual processes or methods embodied in the program are not within the scope of the copyright law.
reinforced by declaring this expression to be a literary work by an addition to the 'literary work' definition (Rep. No. 1476, 94th Congress, 2nd Session 54 [1976]):
..... computer data bases, and computer programs, to the extent that they incorporate authorship in the programmer's expression of original ideas, as distinguished from the ideas themselves .....

This implies a distinguishing line somewhere between expression and idea/process. But where? Without guidance, this Congressional intent is totally useless, as the CONTU commission lamented in 1978:
Where could a meaningful line of demarcation be drawn? Between flow chart and source code? Between source code and object code? ... The Commission believes that none of these is appropriate. The line which must be drawn is between the expression and the idea, between the writing and the process which is described.
Sadly, CONTU's comments are just as useless, little more than word substitutions to Congress' comments.

The software aspects of the 1976 Copyright Act were riddled with confusion. One participant, Congressman Railsback, in a House floor debate in 1976 stated sarcastically, "A good compromise is probably one that satisfies no one, but is acceptable to everyone, and it has been said that this bill is a compromise of compromises.", a double compromise both unsatisfying and unacceptable to anyone.

Quoting a paper by Rick Tache, "Another writer, Litman, argues that courts, in attempting to rely on the legal fiction of legislative intent, are required '... to engage in intolerable contortions and to disregard much of the probative evidence that the legislative history in fact contains.' She provides proof by showing that the court's persistent reliance on older precedent reflects a genuine and understandable confusion about how to interpret the 1976 Act. 'The statute is complicated and its language is often unclear. Courts consult the statute's legislative history for guidance, but find that it compounds their confusion.'"

Any software copyright policies based on the 1976 Act should be repealed.

Dann v. Johnston (Supreme Court - 425 U.S. 219)

Illegitimate progeny of the illegitimate Benson - also ruling that computer programs not patentable, in this case, using an overly broad definition of obviousness to invalid a patent for automatic record-keeping of bank checks.

Parker v. Flook (Supreme Court - 437 U.S. 584)

Illegitimate progeny of the illegitimate Benson also ruling that computer programs are not patentable. "Flook and Diehr: Tweedle Dum and Tweedle Dee in the Supreme Court. ... The issue of patentable subject matter was framed incorrectly in Flook because of the confused analysis in Benson." - see {CHISUM} "In the opinion of the present author, Flook is not a reliable source of intellectual illumination. A thoroughly flawed decision, based on shoddy, superficial analysis, it can only serve to obfuscate." - see {BAUER}.

The Supreme Court's decisions in Benson and Flook affirmed Patent and Trademark Office (PTO) decisions that had denied eligibility. According to Maier, the PTO's theory of unpatentability "had its origins in bureaucratic concerns over workload, rather than in careful theoretical analysis. . . . Worry about workload and backlog motivated the PTO to lead the fight against software patentability." {MAIER} "The acting Commissioner of Patents and Trademarks filed a petition for a writ of certiorari, urging that the decision of the Court of Customs and Patent Appeals will have a debilitating effect on the rapidly expanding computer 'software' industry, and will require him to process thousands of additional patent applications. {from Parker v. Flook}

Thus software patentability was an intellectual mess for decades because some large companies wanted increased hardware sales, and the PTO was too lazy to argue for the resources to do its job well, (all the while software patenting not having a debilitating effect).

In re Sarkar (CCPA - 588 F.2d 1330)

Patent wrongly denied for a geophysical software system citing Benson and Flook.

In re Freeman (CCPA - 573 F.2d 1237)

Along with In re Walter and In re Abele, a partial test, ultimately ruled useless, for what aspects of software are patentable and which are copyrightable (i.e., those aspects not patentable). In 1998, the CAFC commented in State Street about the Freeman-Walter-Abele tests - "has been the source of much confusion" and "has little, if any, applicability to determining the presence of statutory subject matter". Why? The tests were Benson-guilty by association "... the Benson era Freeman-Walter-Abele test -- in other words, analysis as it existed before Diehr and Alappat."

In re Toma (CCPA - 575 F.2d 872)

The PTO Board of Appeals TWICE wrongly supported an examiner's rejection of a software patent application for translating natural language. Both times the Board argued that is what Benson requires. CCPA reversed, saying that's NOT what Benson requires. A good example of just how confusing and poorly worded is the Benson decision.

The CCPA correctly ruled the natural language processing is patentable. But it incorrectly reasoned that a natural language processing algorithm is not a mathematical algorithm, wrongly limiting 'mathematical' to mean 'numerical'. The limitation is one rejected by mathematicians and computer engineers, but one made by courts many times over the years. Interestingly, the PTO Board argued at one point, correctly, that mathematical algorithms include natural language processing.

CONTU (Commission On New Technology Uses of copyright works)

Appointed by Congress to examine software copyright and other problems because of software deficiencies with the 1976 Copyright Act. Sadly the Commission produced a "recommendation appears also to have been based upon a complete misconception of the nature of software ..".

As Samuelson observed about the Commission, "The issue of the copyrightability of machine-readable forms of computer programs was not one on which the Commissioners had any expertise. There were no computer scientists, no software or hardware industry representatives, nor any users of complex software systems on the Commission." And no patent lawyers. Sadly, the Commission consisted of a judge, a copyright lawyer, the Copyright Registrar, three librarians, a novelist/journalist, director of the Consumer's Union, vice presidents of three major publishers, and a general law professor. A selection demonstrating Congress' lack of seriousness in addressing this issue.

Another lawyer writes: "But the recommendation appears also to have been based upon a complete misconception of the nature of software by the majority of the Commissioners." see {BURK}. "Unfortunately, the Commission's conclusion is deficient for not addressing the distinction made in copyright law between the unprotectable idea and the protected expression in a work, such as a flow chart. ... On this point the CONTU Final Report can be criticized. It fails to state with particularity the protectable expression in software and the point at which copyright protection ought to attach. ... Before determining the mode of protection, sui generis or copyright, it is necessary to specify the minimum contribution worthy of protection. The Commission's Final Report provided little assistance for this task." - see {RASKIND}

In their 31 July 1978 Final Report, CONTU had 58 pages devoted to software copyright, in particular the 38 page opinion supporting software copyright (used as the intent of the 1980 Act). However 28 pages were a dissent against software copyright, supported by the few commission members who actually knew anything about computing. The confusion spilled over into a court case in 1982 (Evergreen Consulting v. NCR Comten), where two CONTU commissioners (Miller and Levine) gave expert testimony arguing that copyright law can't be used to protect the structure of a computer program, while another CONTU commissioner (Nimmer) gave the exact opposite testimony.

So subtracting the dissent and the commission's incompetence, and the commission's refusal to criticize Benson (which it cites), and then the many criticisms of the software aspects of CONTU in the following years, and you are left with NOTHING as a basis of federal US, regional NAFTA and global TRIPS software copyright law. Ignore anyone who mentions CONTU without mentioning these many problems.

Data Cash Systems v. JS&A Groups (N.D. Ill. - 480 F.Supp. 1063)

A computer program's copyright is not infringed by a copy of the executable code in a ROM. Relied on the scientific misconceptions of the 1908 White-Smith decision.

In re Phillips (CCPA - 608 F.2d 879)

The PTO Board of Appeals TWICE wrongly supported an examiner's rejection of a software patent application for preparing architectural specifications. Both times the Board argued that is what Benson and Flook require. CCPA reverses, saying that's NOT what Benson and Flook require. A good example of just how confusing and poorly worded are the Benson and Flook decisions.

Copyright Act of 1980

The legal (though not logical) basis for current software copyright, it blindly adopted all of the now discredited software copyright recommendations of the CONTU report. Congress lamely just added a definition of "computer program" to the Copyright Act, forcing the courts to infer from legislative history (i.e., CONTU) that Congress wanted copyright for software. Their inaction was unethical.

Many papers, including Samuelson's "CONTU revisited", Kidwell's "Software and Semiconductors", and Brown's "Eligibility for Copyright Protection" concluded that Congress acted unwisely in protecting software within the framework of copyright. Indeed, 12 years later in the 1992 CA/Altai case, the Second Circuit complained about the Congress' indecision and CONTU's vague recommendations - "Generally we think that copyright registration - with its indiscriminating availability - is not ideally suited to deal with the highly dynamic technology of computer science. Thus far, many of the decisions in this area reflect the court's attempt to fix the proverbial square peg in a round hole. ..... In any event, now that more than 12 years have passed since CONTU issued its final report, the resolution of this specific issue could benefit from further legislative investigation - perhaps a CONTU II." With regards to software, CONTU is CRAPU, and so therefore is the 1980 Copyright Act aspects dealing with software.

In re Walter (CCPA - 618 F.2d 758)

Along with In re Freeman and In re Abele, a partial test, ultimately ruled useless, for what aspects of software are patentable and which are copyrightable (i.e., those aspects not patentable). In 1998, the CAFC commented in State Street about the Freeman-Walter-Abele tests - "has been the source of much confusion" and "has little, if any, applicability to determining the presence of statutory subject matter". Why? The tests were Benson-guilty by association "... the Benson era Freeman-Walter-Abele test -- in other words, analysis as it existed before Diehr and Alappat."

UK Patent Appeal Tribunal: IBM v. UK (1980 FSR 564)

IBM was trying to patent a commodities pricing system. The UK court decided that the invention involved a "special purpose computer" (which is patentable) rather than a programmed "standard computer" (which is not patentable), without explaining any engineering difference between the two embodiments. Similar to the US court's falsely distinguishing between ROM and RAM, such as Diamond v. Bradley. A forerunner of bad semantic by European courts that will argue that software and business methods don't achieve a "technical effect" to be patented, without explaining what a "technical effect" is.

Diamond v. Diehr (Supreme Court - 450 US 175)

Illegitimate progeny of the illegitimate Benson. "Flook and Diehr: Tweedle Dum and Tweedle Dee in the Supreme Court. ... But otherwise [Diehr] purported to confirm Flook, noting for example, that '[a] mathematical formula as such is not accorded the protection of our patent laws' and that 'insignificant post-solution activity will not transform an unpatentable principle into a patentable process'. - see {CHISUM} The court, in the dissent, even noted that Diehr is indistinguishable from Flook.

Diamond v. Bradley (Supreme Court - 450 US 381)

ROM-implementable processes are patentable even if the same process, computer program implemented, would not be patentable. As pointed out by D. Davidson, "This would be silly. The ROM in Bradley could have been replaced in the computer with a volatile RAM memory chip which looks like a ROM; the software could first be read into the RAM, then the process performed." The Bradley decision is indefensible from a factual, i.e., scientific point of view.

In re Abele (CCPA - 684 F.2d 902)

Along with In re Freeman and In re Walter, a partial test, ultimately ruled useless, for what aspects of software are patentable and which are copyrightable (i.e., those aspects not patentable). In 1998, the CAFC commented in State Street about the Freeman-Walter-Abele tests - "has been the source of much confusion" and "has little, if any, applicability to determining the presence of statutory subject matter". Why? The tests were Benson-guilty by association "... the Benson era Freeman-Walter-Abele test -- in other words, analysis as it existed before Diehr and Alappat."

In re Pardo / In re Meyer (CCPA - 684 F.2d 912, 688 F.2d 789)

Pardo was kind of for patenting software, while Meyer was kind of against. As pointed out in {CHISUM}, "What then is the basis for holding that a process for 'executing' 'formulas' with 'data' is not or does not include a mathematical algorithm (Pardo) but that a process for 'initializing' and 'modifying' a 'factor' necessarily is or includes such an algorithm (Meyer)? The basis can only be the stream of judicial confusion from the Benson wellspring."

Apple Computer v. Franklin Computer (3rd Circ. - 714 F.2d 1240)

Upheld copyrightability of computer programs. A "landmark" decision that should be rejected for its reliance on the CONTU report. I often wonder if any of these judges citing CONTU actually read the report and its many criticisms.

Paine Webber v. Merrill Lynch (D.C. Del. - 564 F.Supp 1358)

The court ruled correctly that business methods are patentable. However, there is a statement showing that at least this court still didn't understand mathematics and computing. Their decision notes "Although one may devise a computer algorithm for the Pythagorean theorem, it is the step-by-step process which instructs the computer to solve the theorem which is the algorithm, rather than the theorem itself." This observation is false, even in 1983. In the 1970s, computer scientists developed theorem proving software, often used to prove that algorithms worked properly in all cases. Part of the proof is an algorithm to instruct the computer. Not to be too critical of the court here, since even today most judges and IP lawyers don't realize that theorems can be as executable as algorithms. You figure 30 years of missing the point would get one of them to go ask the mathematicians for some help.

Apple Computer v. Formula Int'l, (C.D.Cal - 562 F.Supp. 775)

Blindly accepts CONTU and the 1980 Act - "Any doubt as to whether the Copyright Act as presently worded protects computer programs of all types, however fixed, is removed by examining the legislative history of the 1980 Act and particularly the work of CONTU.". However, "The greatest defect of legislative history is its illegitimacy" - Judge Scalia, dissent, Conroy v. Aniskoff (507 U.S. 511 [1993]). Also, "Government by unexpressed intent is ... tyrannical. It is the law that governs, ...", Judge Scalia in A Matter of Interpretation (1993). There was no Congressional hearing on the final CONTU report. Also, the court somehow was unable to discover any of the writings severely questioning the 1980 ACT and CONTU, rendering this decision suspect.

Japan Copyright Law Amendment of 1985

In 1984, the Japan Copyright Council released a report recommending that Japan follow the worldwide trend of using copyright law as the primary mechanism of protecting software (as opposed to a sui generis law proposed by MITI in 1984). Although MITI had more power, the Copyright Council's recommendation was followed due to external pressure from the European Economic Community and the United States. The Japan Copyright Law was thus amended in 1985, defining (hopefully with irony) a computer program as "an expression of combined instructions given to a computer so as to make it function and obtain a certain result..", i.e., expressed functionality, and thus inheriting all of the contradictions of American software copyright law.

Semiconductor Chip Protection Act (17 USC 901-914)

A sui generis amendment to the copyright system to protect the "mask work"/topography of integrated circuits. However Congress so narrowed the protection (by allowing reverse engineering), that in its first seven years, the Act was invoked in only one appellate position. The act was forced on and adopted by the Europeans in 1987. By the early 1990s, semiconductor technology developments eliminated any need of the Act to fight piracy, an example of the general uselessness of sui generis systems. See {RAUCH} for a good review of the Act's limitations.

"Six copyright theories for the protection of computer object programs"

Paper by Hardy in the Arizona Law Journal. While rightly rejecting sui generis protection for software, argues that copyright protection "is about as effective a device as one could hope for" for software, ignoring all of the critical analysis to the contrary.

United Kingdom Copyright (Computer Software) Amendment Act

Ambiguous law specifying that computer programs were protected under copyright law as literary works. Meant to be a temporary measure, it was repealed and replaced by the Copyright, Designs and Patent Act 1988. Much like the ambiguity of US law, the 1985 Act did not say outright that a computer program was a species of literary work, but rather presented the position where the UK Copyright Act 1956 was to apply in relation to a computer program as it applied in relation to a literary work.

German Supreme Court Inkasso decision - (Bundesgerichtshof - May 9, 1985)

The Bundesgerichtshof (German supreme court) rules that computer programs are in principle copyrightable without regard to traditional aesthetic conceptions of protectibility, anticipating Whelan's allowing expression to include functional ideas, and anticipating, with a more strict standard, Altai's tough patent-like abstraction-filtration test with a high incremental creativity threshold: "... for [the stages of a program] to be recognized as creative, they must surpass the level of creativity of an average programmer ... the lower boundary of copyrightability begins a considerable further distance away, in the clear surpassing, with respect to the selection, collection, ordering and dividing of the information, of the ability of an average programmer."

Inkasso's "collection, selection, ordering" language is reminiscent of Whelan's SSO "sequence, structure, organization" language. "Surpassing levels of creativity" sounds like unobvious, and "in the clear surpassing" sounds like novelty. So where it escapes the problems of Whelan and Altai, it runs into the patent system.

In 1989, the German Ministry of Justice criticized this Inkasso decision for introducing uncertainty into copyright law in a paper titled "Urheberrechtsschutz von computerprogrammen: kritische Anmerkung zum Bericht der Bundesregierung zur Urheberrechtsnovelle 1985."

SAS Institute v. S&H Computer Systems (M.D. Tenn. - 605 F.Supp. 816)

S&H Computer Systems was found guilty of infringement of the copyright of the non-literal elements of some SAS Institute's software which S&H had ported to a new computer system and had translated into a different programming language. The decision was probably good reasoning within the context of laws at the time, but what gets translated in software ports is the functionality of the source code, which theoretically is the domain of patents, especially if different user interfaces are implemented. SAS should have had patents on the software functionality it was protecting, and sued S&H Computers for patent infringement.

Whelan Associates v. Jaslow Dental Labs (3rd Circ. - 797 F.2d 1222)

Well criticized in too many papers to list here. Example - six years later, in Altai, the 2nd Circuit was very harsh about the Whelan SSO analysis as showing "a flawed understanding of a computer program's method of operation", has a "somewhat outdated appreciation of computer science", and that "... Whelan's general formulation that a program's overall purpose equates with the program's idea is descriptively inadequate".

Whelan contradicts Baker v. Selden by allowing some expression to include functional ideas. "The Whelan test is unsatisfying because of its circularity." and "... both the Whelan test and the Altai test fail to sufficiently distinguish ideas from expression." - see {DIXON}. "The court distorted copyright analysis by treating the protectable interest of the plaintiff as though it were a visual work, when in fact the plaintiff had registered textual material." see {RASKIND}

Introduced the meaningless SSO test for software copyright: Structure, Sequence and Organization. As the Copyright Office rebuked: "The Office will not make registration based solely on a claim of structure, sequence and organization. These terms are ambiguous, and may refer to uncopyrightable ideas, systems, or processes. There, claims described as so-called SSO are unacceptable. [54 Fed.Reg. 13179 (1989)]. SSI is strongly criticized in "Copyright protection of SSO: replete with internal deficiencies and practical dangers." see {KAPPEL}

Whelan also made use of the questionable merger doctrine. "[T]he purpose of function of a utilitarian work would be the work's idea, and everything that is not necessary to that purpose or function would be 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 that purpose; hence, there is expression, not idea." Many severely criticized these comments for defining "idea" far too simplisticly and broadly.

Ignore anyone who mentions Whelan without mentioning these many problems.

Judgments of March 7, 1986 - French Supreme Judicial Court

On this day, the French Cour de Cassation (supreme judicial court) issues three judgments concluding that software is copyrightable, without significant qualifications. One decision introduces the test of 'un apport intellectuel' (intellectual contribution) that requires "a personalized effort going beyond the simple production of an automatic and constraining logic and that this effort [be] materialized in an individualized structure" (the Pachot case). Commentators objected (much like similar objections in the United States to Whelan) that these decisions could be used to extend copyright protection indiscriminately to industrial objects (i.e., objects with 'structure') that are better protected with patents. Fears based on these objections were realized in a April 16, 1991 Cour de Cassation decision lowering the threshold of copyright protection (the case involved graphic elements of a computer program).

Broderbund Software v. Unison World (N.D. Cal - 648 F. Supp. 1127)

Relied on Whelan's reasoning, declaring that "this [c]ourt is persuaded by the reasoning of Whelan", which "stands for the proposition that copyright protection is not limited to the literal aspects of a computer program, but rather that it extends to the overall structure of a program". A decision as bad as Whelan - "The Broderbund court seems to have eviscerated the First Amendment limitation by abandoning any meaningful attempt to distinguish ideas from expression." - see {DIXON}

"The bundle of rights suited to new technology" (Stern)

Law review article. Proposes a sui generis registration system for the noncode aspects of software - a registration without a determination as to the technical merits or validity of the work unless and until it is challenged in an infringement action, protection somewhere between that of a patent and copyright, for five to twenty years. A de jure system later shown to be impractical by the de facto registration process that patent offices around the world have reduced to under examination workloads beyond their means. His definition of "code" is vague, and then cites but does not challenge the flawed Apple/Formula decision or the decision's reliance on the CONTU.

Computer Edge Pty. v. Apple Computer (Australian High Court)

A decision with a confusing history. In 1983, the Federal Court of Australia ruled that computer programs were not literary works and therefore not protectable by copyright. In 1984, on appeal, the Full Court of the Federal Court reversed, holding that both source code and object code were copyrightable. On further appeal, the High Court held that while source code programs were literary works to be protected by copyright, the programs in object code were not literary works and therefore not protectable. A bad decision in the spirit of the 1979 Data Cash v. JS&A Groups decision, where a computer program's copyright is not infringed by a copy of the executable code in a ROM.

Cox v Riley (UK 83 Cr App R 54)

A to-be-reversed UK Court of Appeal ruling that computer programs stored on magnetic media are not tangible property (to be protected, for example by copyright). Reversed five years later in the case of R v. Whiteley (UK Cr App R 381), which ruled that it did not matter if the damage itself was not tangible, providing tangible property had been damaged.

Australian Patent Office guidelines for computer inventions

In March 1986, the Australian Patent Office publishes a new set of guidelines: Guidelines for Considering the Patentability of Computer Program Related Inventions. These guidelines incorporated the Freeman-Walter-Abele test. Challenges in courts over what exactly was so patentable in the 1986 guidelines (reflecting similar confusion in the US) led to new APO guidelines in August 1992 with a new test for software - "does the invention claimed involve the production of some commercially useful effect?", the "useful" part of the American Alappat and State Street decisions.

US Office of Technology Assessment report - "Intellectual Property Rights in an Age of Electronics and Information"

Government report (by people who actually understand technology) that criticizes recent copyright decisions allowing protection of non-literal aspects of software. They conclude with:
In theory, none of these rulings is permitted under traditional copyright principles. This is not because the courts have misinterpreted copyright law, but because copyright law cannot be successfully applied to computer programs. Unlike artistic or factual works, which are subject to the "abstractions" test between idea and expression, copyright in computer software is an either/or choice . . . The "clear distinction" made in Baker v. Selden, 'between the book, as such, and the art it is intended to illustrate', collapses in a computer program, because the program embodies both "the book" and "the art".

Plains Cotton Coop. v. Goodpasture Computer Service (5th Circ - 807 F.2d 1256)

While rightly deciding that the overall structure of a computer program is not copyrightable, the court, in wrongly relying upon an earlier decision (Synercom), falsely states that input formats to a computer program are "a level of computer software design more specific than functional design and more general than line-by-line program design".

Petition for a Writ of Certiorari denied by Supreme Court for Whelan

The Supreme Court refuses to examine the Whelan decision.

EC Semiconductor Products (Protection of Topography) Regulations

Under pressure from the United States, the EC introduces a sui generis system of protection for the layout of an integrated circuit, based on the US' 1984 Semiconductor Chip Protection Act.

Microsoft v. Shuuwa Trading (D.C. Chisai)

A Japanese district court concludes wrongly, in the mode of the Whelan philosophy, that the composition of procedures and data structures are nonliteral elements deserving of copyright protection.

Digital Communications v. Softklone Distributing (N.D.Ga - 659 F.Supp. 449)

User interface copyright infringement case. A confusing decision based on an earlier confusing decision, Synercom Technology v. University Computing. In an article mentioning Softklone, Teter writes, "In holding for the plaintiff, the court confessed that although it it found Synercom 'somewhat confusing', it considered the Synercom hypothetical to mandate a simple metaphysical rule: 'idea' is the process or manner by which the status screen, like the car, operates and the 'expression' is the method by which the idea is communicated to the user. Needless to say, Softklone does little to clarify the status of dynamic merger."

Pearl Systems v. Competition Electronics (S.D.Fla - 8 USPQ.2d 1520)

"... example of misapplied copyright relief granting patent-like protection without regard to the novelty and nonobviousness requirements of the patent system." Extended software copyright protection to patentable functionality. Also relied on Whelan.

Healthcare Affiliated Services v. Lippany (W.D. PA - 701 F.Supp. 1142)

Another bad Whelan-like decision, adding a new level of abstraction for a copyrightable aspect of software to the Whelan SSO levels, "methodologies". Ineptly, "methodologies" wasn't precisely defined (not that it could be without grounding it in "process" and thereby triggering a 102b prohibition) in the decision.

"Reflections on the state of American software copyright law and the perils of teaching it", Pamela Samuelson, Columbia-VLA J.L & Arts 1988, 61

"Teaching software copyright law in the United States is at present a perilous endeavor .. When a whole field of law is a welter of confusion and contradiction, it is not small challenge to teach the law as it truly is and keep students' attention and respect." I disagree. It is very easy to teach that software does not fall in the domain of copyright law, a respectable position with no confusion and no contradiction.

!UK Copyright, Designs and Patents Act 1988

Introduces the same ill-defined assertion as found in US copyright law of 1976 and Berne convention - that computer programs are literary works, without explaining why or how. Section 3(1) has this as "a computer program", "a table or compilation", and "preparatory design material for a computer program", i.e. mostly software, data structure, and software enginering data structure ideas prohibited from copyright by the 1991 EC Directive. Worse, the 1988 Act treats computer programs as an inferior class of literary works, denying the author of a computer program the protection of moral rights in his work - Section 79(2). Section 21.3.ab extends copyright protection to "an arrangement or altered version of the program or a translation of it", without defining the concept of an "arrangement" of a computer program and without recognizing that often translations of computer programs are functional implementations of an algorithm in one language in another language.

Ahitov v. Harpaz (Israel)

A Tel Aviv district court wrongly holds that creating a new computer program based on the functional specifications and system design of another computer program violates the copyright held by the author of the original program.

Johnson Controls v. Phoenix Control Systems (9th Circ. - 886 F.2d 1173)

Another bad Whelan-like SSO decision, though not citing Whelan, held that "[w]hether the non-literal components of a program, including the structure, sequence and organization and user interface, are protected depends on whether, on the particular facts of each case, the component in question qualifies as an expression of an idea, or an idea itself."

Manufacturers Technologies v. CAMS Inc. (D. Conn - 706 F. Supp. 984)

Decision wrongly ruling GUIs are copyrightable - "plaintiff's screen displays . . . merit protection as copyrightable expression". A bad decision - "... example of misapplied copyright relief granting patent-like protection without regard to the novelty and nonobviousness requirements of the patent system." Extended software copyright protection to patentable functionality.

FSF's Copyleft / GPL - General Public License

Both the Free Software Foundation's Copyleft and General Public License (GPL) used by the open source software communities rely on software copyright. For example, you see language like the following - "... either the Program or any derivative work under copyright law ... " in the GPL. If software copyright is very weak, so too is copyleft and the GPL. A good critique of GPL's use of copyright to achieve excessive control is in {HAYNES} - "In the software arts, including e-commerce, the system is unbalanced by the tendency of copyright law to isolate technology in pockets controlled by the original author, and thus to act as the gravity for 'black holes' of innovation. ... However, even the open source movement exploits and is hindered by copyright." The FSFers and GPLers have been misled on the usefulness of copyright as an enforcement mechanism for copyleft and GPL.

"Patentability of dedicated information processors and infringement protection of inventions that use them" (Dukarich - Jurimetrics Journal)

Argues contemptuously that implementing a well-defined process in a programming language is always obvious and therefore nonpatentable. Since the process can be programmed in VHDL, I guess he is also arguing that hardware should be nonpatentable.

Lotus Dev. v. Paperback Software (D. Mass - 740 F. Supp. 37)

A long and ultimately confusing opinion, and also endorsed Whelan's protection of SSO. The opinion was eventually totally rejected by a higher court.

"Copyright at the School of Patent", John Shepard Wiley Jr., U. Chi. L. Rev. 119

In many software copyright cases, the courts rely upon the merger doctrine to objectively find some software expression to protect distinct from the software's ideas. In this paper, Wiley argues there is no such objectivity, and that the merger doctrine's arbitrary nature is such that it should not be used (for example, as in Altai in 1992, and Softel in 1997). Some excerpts from his paper:

3. The manipulability of merger.

Later opinions add material for this critique by embroidering the idea/expression dichotomy with the thread of the so-called merger doctrine. Merger doctrine holds that copyright protection lapses for a work in which idea and expression "merge" or become "inseparable." We should become suspicious upon hearing that it is possible for two supposedly dichotomous concepts to merge into inseparable unity. The doctrine of merger announces this phenomenon, however, whenever "there are no or few other ways of expressing a particular idea" apart from the plaintiff's particular expression. In such instances, courts claim to break with their normal practice by giving later authors freedom to copy the first author's expression. The rationale is that this freedom is necessary to liberate for copying the first author's idea -- which axiomatically can never be subject to copyright protection.

The case commonly cited as the doctrine's source and its classic illustration, even though it failed to use the word "merger", is Morrissey versus Proctor & Gamble Co. Morrissey sued Proctor & Gamble for copying the rules for a promotional sweepstakes contest based on participants' Social Security numbers. Notwithstanding the nearly verbatim copying, the court worried that "to permit copyrighting would mean that a party . . . by copyrighting a mere handful of forms, could exhaust all possibilities of future use of the substance". The court thus excused Proctor & Gamble's copying because Morrissey's idea (or substance) and expression had merged.

The Morrissey result probably makes sense. But indeterminacy overwhelms its logic of merger. Before a court can decide whether "there are no or few other ways of expressing" plaintiff's idea, it must define the plaintiff's idea. No unique abstract description or idea is self-evidently appropriate; instead there exists, as Judge Hand pointed out, "a great number of patterns of increasing generality [that] will fit equally well, as more and more of the [detail] is left out". As usual, the crucial doctrine remains mute about how far up the scale the judge should proceed before testing the number of alternative expressions. Without a scale, it is impossible to say. Nonetheless, cases applying the merger doctrine inevitably do say, settling on an implicit specification of idea that confines alternative means of expression to a relatively small number. Also inevitably, however, a variety of alternative definitions remains both conceivable and consistent with the malleable doctrinal terms. The merger doctrine's use of idea and expression thus is characteristically open-ended -- useful for announcing a decision, but not for generating or justifying one.

To demonstrate a critique that applies to every case invoking or rejecting the merger doctrine, observe first that the Morrissey court implicitly defined the idea as "advertising using a sweepstakes based on social security numbers". Defining idea so narrowly does leave a relatively small number of alternative expressions. (Actually, the number of equivalent rephrasings probably runs to the hundreds or thousands, but this quibble is at once digressive and fantastically tedious to verify.) We can multiply these alternative possibilities to a countlessly large number and thus negate merger simply by choosing a more abstract definition of the idea. For instance, it is equally accurate to define the idea of Morrissey as "advertising using a sweepstakes based on a number". One can generate contest numbers in an endless number of different ways: phone numbers, birthdays, addresses, guesses about the daily volume of the New York Stock Exchange or the point spread in some athletic contest, or eight digit numbers picked at random. In doctrinal terms, merger dissolves and Morrissey wins. To the same end, we could define Morrissey's idea simply as "advertising using a sweepstakes", or even more simply as just "advertising". The merger doctrine contains no logical rule for selecting the decisive level of abstraction.

As I have already said, Morrissey's outcome probably makes sense, as do the outcomes in Nichols and Sheldon. But the doctrine that justifies these results is inconclusive and ultimately arbitrary. Every mention of idea and expression states legal conclusions while doing nothing to justify them. Every utterance of the terms should prompt lawyers and judges to ask, "Why does copyright law allow defendants to copy this thing labeled an 'idea'? How do we know that plaintiffs cannot appropriate some other thing called 'expression'?" To these basic questions, the doctrine of idea and expression can offer no answers. The kindred merger doctrine merely buries more deeply the questions it begs.

European Union Directive of 14 May 1991 on software copyright (91/250/EEC)
Well critiqued in "Criticism of the E.U. Directive protecting computer software". See {SHAPOSKA}

EC Directive on Legal Protection of Computer Programs (91/250/EEC)

The Europeans extend copyright protection to software with the same ill-defined legal language that appeared in the earlier US 1976Act/CONTU/1980Act. Article 1.1 invokes the Berne treaty to declare computer programs as literary works. But Article 1.2 uses the language of 17 USC 102b that the ideas of a computer program are not copyrighted, only the expression. And much like the US, the EC provides no guidance on how to reconcile the two clauses (or even showing that they are reconcilable, which the US courts have failed to achieve). Thus the EC does little more than introduce American confusion into European software copyright law, made worse as the members of the EC enacted the EC directive into national laws. Well critiqued in "Criticism of the E.U. Directive protecting computer software". See {SHAPOSKA}

Microsense Systems Ltd v. Control Systems Technology Ltd (UK unreported 17 July 1991 Chancery Division)

Judge rules that a list of mnemonics for programming a traffic control system, were protectable by copyright. The US courts decided correctly in the 1995 Mitel v. Iqtel case that instruction sets were not copyrightable, being merely a procedure, process, system and method of operation.

Re Gale's Patent Application (UK RPC 305)

A UK Court of Appeal ruling against software patents. A patent was sought for a computer program in ROM for calculating square roots. The Court ruled the computer program was unpatentable since it was entirely software based and the fact that the program was stored on an integrated circuit did not alter this finding. Obviously the court was unaware of FPGAs, which allow programs to be directly mapped into patentable hardware, making ill-defined the phrase "entirely software based".

"Copyright protection for artificial intelligence systems", Goldberg and Carson, J. Copyright Society of USA, Fall 1991, 57

Argue wrongly that expert systems deserve copyright protection. "Our thesis is that AI systems face - and should face - no more difficulty than traditional computer programs in obtaining copyright protection. Since the world copyright community is in general agreement that computer programs are entitled to copyright, AI systems should enjoy the same status. .... But without the exclusive rights of copyright, the production of AI systems will be impaired." If there is anything more process and procedure than the rules of AI/expert systems, I have no idea what it is.

The authors also cite with approval the ill-defined 1976 Copyright Act and CONTU writings, introducing their own ill-defined bad science, writing "... We believe an AI system also contains expression of the process and method, and that - as with computer programs generally - such expression is copyrightable." The article is mostly devoid of mention of using patents to protect AI systems, even though over 200 patents in the 1985 to 1990 time period mention expert systems and artificial intelligence.

Computer Associates v. Altai (2d Circ. - 982 F.2d 693)

The decision creates the abstraction-filtration-comparison test to deal with the problems of trying to find any expression in software. The decision is severely criticized in the years that follow to the extent that seven years later in 1999, a Second Circuit judge, Judge Jon Newman, questions the test "... and I am not sure we ought to have something called an 'abstraction-filtration-comparison test'". Cites CONTU without citing the many criticisms of CONTU, while not citing the 1986 OTA Report.

Forced by the idiocy of CONTU and the 1980 Copyright Act, the Second Circuit turns software copyright into a patent system without claims, the so-called abstraction-filtration-comparison test. Loosely, software copyright infringement analysis became a bottom-up Markman hearing without claims (abstraction), a novelty analysis in light of scenes-a-faire and public domain prior art (filtration), finishing with an obvious analysis in light of the "golden nuggets" that remain (comparison). Indeed the court almost concedes this in a comment - "... patent registration, with its exacting up-front novelty and nonobviousness requirements, might be the more appropriate rubric of protection for intellectual property of this kind.". Nice wisecrack - 'patent registration'.

Indeed the district court from which this was appealed, plus another commentator cited by the Second Circuit, had argued that all of this should be thrown over into the patent world with its formal 102 and 103 tests in light of explicit structural claims. And I suspect the Second Circuit would have loved to transfer all of these software IP matters over to the patent system. But it couldn't, because it had to follow the Congressional intent of the 1980 Copyright Act (based on the poorly reasoned 1978 CONTU report).

Even worse, the Altai court based its three-part test on three-part tests proposed by Nimmer in his treatise (the Nimmer of the discredited CONTU) and Judge Keeton in his Lotus v. Paperback decision (which was later overturned).

And because the court was forced to maintain the fiction of handling software with copyright, Altai's reasoning contributed to the confusion. "... both the Whelan test and the Altai test fail to sufficiently distinguish ideas from expression." - see {DIXON}. "However the failure of the court to be explicit in this regard [only code is protected against literal copying] has left most later courts still floundering, even where they purport to follow the Computer Associates analysis. The basic question that Computer Associates failed to address is how any nonliteral program aspect that somehow survives the filtering procedure can be deemed expressive." - see {KARJ2}. The Altai test is "neither useable nor comprehensible ... a legal Chernobyl" - see {CLAPES}.

Indeed, there are software tools published before the Altai decision that automatically perform bottom-up abstraction on source code - "This paper describes the development of a tool that abstracts formal specifications from program code, using formal methods and object oriented techniques" (from a 1991 IEEE conference). The resulting abstraction data structures from such tools are inherently patentable data structures (heck, the tools can be adapted to write patent claims), and not copyrightable expression. Yet there is no reflection in Altai and other decisions that the process of abstraction, and the output of the abstraction process (a bottom-up analysis), are in reality functional data structures and processes and therefore the domain of the patent system. That the courts and copyright lawyers are too lazy to learn about such software tools is no justification for treating abstraction as non-functional.

The abstraction-step is an illegitimate software test, a pale imitation of formal software and engineering and patent claims drafting. Without it, Altai is useless. That the district and circuit courts keep on issuing software decisions without researching software engineering techniques is unprofessional. It is a contempt for another profession.

Many copyright cases since then have cited Altai. The courts should be showing more leadership when forced to rely on Altai by explicitly putting into their decisions a charge to Congress to remove software from the copyright statutes. Because that is all that Altai and its progeny are - a patent registration system without patent claims, but rather some arbitrary bottoms-up structural analysis done by judges.

Patent lawyers should be more openly critical of Altai for its incursion into their domain, and software engineers for the courts' continual refusal to learn much about software engineering and the many tools developed.

Ignore anyone who says we can live with Altai.

"Computer Associates v. Altai: rationality prevails" (David Bender)

August 1992 law review article praising Altai. Of course Altai is rational - rational plagiarism - since it borrowed from the rational patent system. But Altai doesn't deserve praise for borrowing patent practices without patent claims.

CMAX/Cleveland v UCR (M.D. Ga 804 F.Supp. 337)

Allows software copyrighting of what the court explicitly labeled as the "system" and "system design" aspects of a computer program's user interface, even though such aspects are functionally patentable. Well critiqued in {KARJ2}.

"Defining computer program parts under Learned Hand's abstractions test in software copyright infringement cases" (John Ogilvie)

Paper by an IP lawyer with a master's degree in computer science, supporting software copyright. Tries not very well to come up with a practical implementation of Hand's thoughts on abstraction in Nichols (he refers to Hand's thoughts as a test, even though Hand himself said years later it wasn't a test). Like the Altai court, the author ignores much software research into tools for automatically abstracting software (preferable to software-naive judges doing reverse engineering).

Inaccurately states that source code must be translated once per hardware type to produce object code (emulators can run one binary on another platform, which is a translation, but in a sense different from how he is using the word, and should have been noted), nor does he mention that there are specialty processors that directly execute source code (using on-chip interpreters, which is not the same as software interpreters). Makes repeated use of "abstract" and "abstraction" without either defining the words or noting that they are highly ambiguous (a deficiency of State Street as well). Fortunately, this paper is only cited in one case, Gates Rubber (see below).

"Current status of patent protection for computer software" (Capes)

Article in JPTOS, January 1992. Tries to explain how to patent computer software. Mentions Benson without criticizing it, and suggests that "The decision in Paine Webber is probably incorrect", a guess later proved wrong by the State Street and ATT/Excel decisions.

Autodesk v. Dyason (Australian High Court)

"The High Court finds that Dyason and others infringed the copyright of AutoCAD by producing and selling a "dongle" replacement. Apart from appearing to endorse Whelan v Jaslow, the really scary thing is that copyright infringement was found because the "dongle" replacement reproduced a 121 bit look up table and this was a substantial part. OK, I tell a lie, the really scary bit is that neither of the parties had actually argued this so there is Autodesk v Dyason (No 2) in which the High Court by a 3 - 2 majority decided they had given Dyason a fair hearing." - as described to me by an Austrlian patent lawyer.

NAFTA and TRIPS (sections of software copyright)

Under US dominance, both the NAFTA and TRIPS agreements provide that computer programs are copyrightable literary works, with Congress taking the idiocy of 1976Act/CONTU/1980Act and jamming it into these two new IP treaties. For example, the TRIPS Agreement, Section 1, Article 10 states: "Computer programs, whether in source or object code, shall be protected as literary works under the Berne Convention (1971)." The Europeans should have fought the Americans harder over the issue of making software copyrightable in TRIPS.

Autoskill v. National Educational Support (10th Circ. - 994 F.2d 1476)

Another bad software copyright decision in the spirit of Benson. "The court did not, however, explain why the functional choice of which sub skills to test for constituted copyrightable expression. As such, the Tenth Circuit's application of its test, like the Whelan test, provided copyright holders with a very broad scope of copyright protection for software.". see {DIXON} "An even more questionable purported application of Computer Associates [than CMAX] is found in Autoskill ...... [it] is in fact much more closely aligned with Whelan and its progeny". see {KARJ2}.

MAI Systems Corporation v. Peak Computer (9th Circ. - 991 F.2d 511)

Ruled that loading a computer program into RAM involved making an infringing copy (which was being done in this case for maintenance purposes). A ruling that was in conflict with a 1976 House Report, (H.R. Rep. No. 1476, 94th Congress, 2d session, 52-53, 62 - where Congressional intent is that RAM loads are unfixed, evanescent 'images', not copies) and the general nature of Section 117 of the 1980 Copyright Act. In 1998, Congress modified Section 117 to exempt from infringement the very type of software loading done in MAI Systems. MAI Systems has been criticized in many law review articles.

In this case, the 9th Circuit affirmed a California district court decision. An ex-employee of MAI went to work for Peak, so MAI sued both of them. Jessica Litman, in her 2001 book, 'Digital Copyright', has a cutting bit of sarcasm about the district court judge: "The judge apparently decided that this guy was scum, and he handed down one of those unfortunate opinions where the plaintiff prevails on every single claim and the defendant's every argument is wrong."

Gates Rubber v. Bando Chemical Industries (10th Circ. - 9 F.3d 823)

A failed attempt to improve on Computer Associates v. Altai. Relied on the Whelan SSO, CA/Altai and Autoskill cases (which have been discredited for their incoherence), commenting "We find in these and other cases .. that there has begun to be developed a coherent approach to the protectability analysis." Also relied on Lotus Dev. v. Paperback Software (D. Mass - 740 F. Supp. 37), which itself was later reversed.

"Like Computer Associates, the court did not fully realize the implications of its approach for the protection of nonliteral program elements; it certainly did not expressively limit protection to literal code (or even derivatives of literal code) and even suggested, without supplying any standard, that some aspects of structure may be 'expressive'." - see {KARJ2}

GNU Business Information Systems v. Social Secretary Ltd (N.D. Ill)

Another case awarding copyright status to functionality. Here an arrangement (i.e., functionally structured) public domain subroutines may be protectible, as may be the selection and arrangement of certain unprotected elements.

"Copyright protection for computer programs, databases and computer generated works: is anything new since CONTU?" by Arthur Miller, 1993 Harvard Law Review, 977

Prof. Miller, one of the CONTU commission members, unsuccessfully tries to defend CONTU and its aftermath. Makes the argument (as have others who haven't realized that Mel Nimmer in his treatise omitted [and is still missing from Nimmer on Copyright] one major body of science that radically alters copyright) that programmers are "isolated artists" who need the protection that artists have - copyright. Well critiqued in {KARJ1}.

Rosenthal v. The Israeli Patent Registrar (D.C. Tel Aviv - no. 598.93) United Technologies v. The Israeli Patent Registrar (D.C. Jerusalem - 23/94)

The Israeli patent office denies a patent on a numerical control algorithm filed as a general software algorithm, upheld by the District Court on mental steps ground (an argument eventually rejected in the United States) that the patent included only calculation steps without any effect on physical matter. The District Court decision was delivered on 11/21/96. One year later in 1994, the Patent Registrar rejected a United Technologies patent for software to control fuel composition in a helicopter, arguing that the Rosenthal decision prohibits protection for software-only inventions. On appeal, the Jerusalem District Court overturns, citing the 1981 U.S. Supreme Court decision, Diamond v. Diehr. After that, Israeli software patent policies tended to follow the U.S.

Denmark computer program copyright act
German computer program copyright act

Denmark and Germany implement the EC 91/250/EEC directive on software copyright, the directive lacking clarity on the resolving the idea/expression dichotomy for software. Also Ireland, Spain and Greece do similar implementations in 1993.

1993: John Richardson Computers v. Flanders (UK FSR 497)

Judge rules that a pharmacy software package had its copyright infringed, relying on the mostly useless US Altai decision: "There is thus nothing in any English decision which conflicts with the general approach adopted in the Computer Associates case ...". The judge based his assessment on looking at just the program's display screens.

Ibcos Computers v. Barclays Mercantile Highland (UK FSR 275)

Judge rules that a database structure is protected by copyright, "... there may well be a considerable degree of skill involved in setting up the data division of a program .. and its construction may well involve enough skill, labour, and I add, judgment, for it to be considered a substantial part of the program as a whole." The decision fails to consider the functional idea aspects of database structure that are prohibited from copyright by the 1991 EC directive. Another UK judge had more correctly ruled to the contrary two years earlier in Total Information Processing Systems v. Daman (see below).

Triad Systems v. Southeastern Express (N.D. Cal. - 31 USPQ.2d 1239)

Ruled that loading a computer program into RAM involved making an infringing copy (which was being done in this case for maintenance purposes). This decision was based on the much criticized 1993 MAI Systems case. In 1998, Congress modified Section 117 to exempt from infringement the very type of software loading done in Triad Systems.

Advanced Computer Systems v. MAI Systems (E.D. Va. - 854 F.Supp. 356)

Ruled that loading a computer program into RAM involved making an infringing copy (which was being done in this case for maintenance purposes). This decision was based on the much criticized 1993 MAI Systems case. In 1998, Congress modified Section 117 to exempt from infringement the very type of software loading done in Advanced Computer Systems.

Kepner-Tregoe v. Leadership Software (5th Circ. - 12 F.3d 527)

Software copyright decision adopting Altai and therefore illegitimate whatever the conclusion. Infringement found in non-literal sequence and organization. The court stated "to determine the scope of copyright protection in a close case, a court may have to filter out ideas, processes, facts, idea-expression mergers, and other unprotectable elements of those materials." Gee, sounds like determining the scope of patent claims to me.

Engineering Dynamics v. Structural Software (5th Circ. - 26 F.3d 1335)

Software copyright decision adopting Altai and therefore illegitimate. Gives broad copyright protection to functional program interfaces (in this case, the input format of a computer program).

Cognotec Services v. Morgan Guarantee of NY (S.D.N.Y. - 862 F.Supp. 45)

Software copyright decision adopting Altai and therefore illegitimate. The court called the Altai test ".. necessarily fact determinative and requires detailed comparison of computer programs.". Sounds like claim drafting and infringement analysis to me.

In re Trovato (CAFC - 42 F.3d 1376)

Patent case counter to the growing trend of software patentability. "Although the Trovato decision should be criticized for showing so little a grasp of technology at issue that it reflects badly on the Federal Circuit, more interesting insights can be gleaned from the reasoning of the decision. ... But the lack of an explanation as to why a programmed computer implementation is patentable subject matter in Warmerdam but not in Trovato, again, is perhaps not merely another technical lapse or oversight in judicial craftmanship. ... Trovato makes no factual sense because its reasoning fails to comprehend what is elementary to an electrical engineer - programming a microprocessor is just another way of making circuitry." see {GOODMAN}

In re Warmerdam (CAFC - 33 F.3d 1354)

Another patent case, with some potential, but with some problems. "The difficulty is that there is no clear agreement as to what is a 'mathematical algorithm', which makes dicey the determination of whether the claim as a whole is no more than that." - a quote from the decision. There is much agreement, just go ask mathematicians to supply one. There are entire conferences and journals on the theory of algorithms in their many forms, all well-defined and easily applicable to patent law. It is an insult to mathematics for such statements to be asserted in the 1990s by any court, a contempt to not ask for help.

"Manifesto concerning legal protection of computer programs" (Samuelson ..)

While the paper correctly attacks software copyright, it also attacks software patenting, relying on the fiction that that software has 'behavior' ignoring the fact that dictionaries define 'behavior' as 'how a machine functions' - functions - process - patents.

Proposes a de jure sui generis registration system for software (while not conceding the failure of de facto registration systems for software [i.e., the world's patent offices], similar to the registration system suggested by Stern in 1986.

Well critiqued in "Comments on a manifesto concerning the legal protection of computer programs". see {GORMA} Also well critiqued in "Four reasons and a paradox: the manifest superiority of copyright over sui generis protection on computer software".

"Four reasons and a paradox: the manifest superiority of copyright over sui generis protection on computer software" (Ginsburg)

While a nice critique of sui generis protection for software, it ignores the functionality of software for which copyright offers no protection, and offers no rationale for distinguishing patent and copyright subject matter. She refers without much criticism to the very poorly drafted aspects of the 1976 and 1980 Copyright Acts that dealt with software. She supports various Courts' acceptance and use of the discredited SSO test of Whelan.

Belgian computer program copyright act
French computer program copyright act (10 May 1994)

Belgium and France implement the EC 91/250/EEC directive on software copyright, the directive lacking clarity on the resolving the idea/expression dichotomy for software. The Netherlands do something similar in their laws reflecting EC 91/250/EEC.

"Intellectual Property and the National Information Infrastructure: Preliminary draft of the report of the working group on IP rights"

Cites CONTU/1980Act without criticism and ignores twenty years of software copyright controversies. Working group chaired by the copyright-lawyer Commissioner of the Patent Office (showing just how screwed up IP politics was becoming).

Amicus Curiae Brief of DEC/GatesRubber/Intel/Xerox supporting Lotus

Relies on the discredited Whelan, Gates/Bando and CA/Altai cases. Unethically cites CONTU without mentioning the many flaws of CONTU, while not citing the opposing 1986 OTA Report.

Amicus Curiae Brief of AIPLA supporting neither Lotus nor Borland

Section II.B is titled "The First Circuit Approach Ignores the Intent of Congress as Reflected in the Legislative History" and cites as an authority the 1976 House Report's ill-defined comments on "the basic dichotomy between expression and idea" remaining unchanged, and similar confusing comments from the CONTU Final Report. The AIPLA unethically leaves out the many previous legal criticisms about both the House Report and the CONTU Final Report. Section D is titled "The First Circuit Approach Misinterprets Baker v. Selden and the Merger Doctrine", but doesn't cite any criticisms of the merger doctrine, especially as applied to computer software (see, for example, the Wiley paper referenced above). One hopes AIPLA members prepare better briefs for the real clients.

Productivity Software Intl. v. Healthcare Technologies (S.D.N.Y.)

Software copyright decision adopting Altai and therefore illegitimate, though it did rule that non-literal elements of some software were not infringed under copyright.

Control Data Systems v. Infoware (D. Minn - 903 F.Supp. 1316)

Software copyright decision adopting Altai and therefore illegitimate.

"Computers, copyright and substantial similarity: the test reconsidered", by Jeffrey Coulter in J. of Computer and Information Law, 1995, 47.

To quote the author "The Altai test is indeed well-considered." and "Altai is to be praised for its synthesis of previous case law, and for the orderly procedure with which it undertakes analysis of substantial similarity." Of course Altai is well-considered and a good synthesis of case law - patent case law though, something that escaped the author. Mentions the 1976 Copyright Act, CONTU and Whelan without criticism.

Harbor Software v. Applied Systems (S.D.N.Y - 925 F.Supp. 1042)

"One decision, Harbor Software v. Applied Systems, analyzes several program structure and interface aspects and, basically, finds copyright protection for all of those for which alternative choices were available, including the selection and arrangement of data. This overly simple analysis fails to recognize that Sect. 102b and Baker v. Selden go well beyond merger to the heart of the distinction between patent and copyright subject matter." - see {KARJ2} The opinion refers to many exhibits only by their title, so its holdings are a little unclear. It found some elements protectible (such as specific code, database fields, "data flow relationships"), and some not (such as "structure and data flow", file locking features, certain flow charts, algorithms, and other methods of operation).

The funny part of this decision is that it ruled that the Altai patent-plagiarizing abstraction-filtration test is for the court to apply, not a jury, citing ..... Markman, the decision ruling patent claims structure and scope are to be determined by judges.

Bateman v. Mnemonics (11th Circ. - 79 F.3d 1352)

Software copyright decision adopting Altai and therefore illegitimate. At least the court recognized the existence of the patent system, though in a half-hearted way, at footnote 21 - "It is particularly important to exclude methods of operation and processes from the scope of copyright in computer programs because much of the contents of computer programs is patentable. Were we to permit an author to claim copyright protection for those elements of the work that should be the province of patent law, we would be undermining the competitive principles that are fundamental to the patent system." The court, in not understanding that all elements of software are the province of patent law, is indeed being undermining.

MiTek Holdings v. Arce Engineering (11th Circ. - 89 F.3d 1548)

Software copyright decision adopting Altai and therefore illegitimate. Actually worse - skipped the abstraction test.

"Copyright and computer programs: it's all in the definition (Patry, Cardozo Arts & Entertainment Journal)

Strongly supports the copyrightability of software. Well critiqued in {KARJ2}.

EC Directive on Legal Protection of Databases (96/9/EEC)

Extending the confusion of its 1991 Directive on Legal Protection of Computer Programs, the EC in Article 3.1 rules that the structure (selection/arrangement) is protectable by copyright, but the protection does not extend to the contents (which can be separably copyrighted). But the EC doesn't explains how to reconcile the software ideas inherent in the structure of a database with its 1991 Directive that said that software ideas are not copyrightable. In many instances before 1996, the US Patent Office had been issuing patents on database structures, an approach nowhere reflected in the EC Directive.

Softel v. Dragon Medical & Scientific Comm. (2d Circ. - 118 F.3d 955)

Another bad decision sustaining software copyright, weakening Altai's abstraction-filtration-comparison test by allowing combinations of unprotectable software to be protectable. Again another patent concept (combinations of existing technology) - combination implies structure and process.

"The Second Circuit's recent decision in Softel rather dramatically shows the difficulty courts have in keeping fundamental principles in mind trying to apply the Altai filters. ... The beguiling but wholly inapt analogy to the esthetic core of copyright, such as paintings and novels, causes court like that in Softel to lose sight of the basic policy goals for protecting technological works like computer programs under copyright law and runs the risk of that they will ultimately obliterate the boundary between patent and copyright. ... The Softel opinion also appears to make the mistake of equating expression at structural program levels with the mere absence of merger." see {KARJ2}

Mitel v. Iqtel (10th Circ. - 124 F.3d 1366)

Software copyright decision adopting Altai and therefore illegitimate. "Although we conclude that although an element of work may be characterized as a method of operation, that element may nevertheless contain expression that is eligible for copyright protection....", making it clear that the court never bothered to read any of the journals or conference proceedings on user interfaces, user interface design tools, and human-computer interactions, with many papers to the contrary. The district court (D. Colo. - 1995 - 896 F.Supp. 1050) held that the telephone input command codes were an uncopyrightable "procedure, process, system, and method of operation".

"Software law-lessons from America: filtration of functionality from software copyright" by P.Stone in Computer Law and Security Report, Jan.-Feb. 1997, vol.13, (no.1):15-21.

The author discusses the disagreement among English judges as to whether the US approach towards objective similarity between computer programs (involving the abstraction, filtration comparison test first announced in the 1992 US decision in Computer Associates v Altai) offers a useful model to be followed in the UK. Apparently the English as well haven't figured out that Altai is a patent system without claims, which they don't need as a model (nor does anyone) since they already have a patent system.

Powerflex Services v. Data Access Corp. (490 Federal Court of Australia)

Powerflex created a computer program highly compatible with Data Access Corporation's fourth-generation language DataFlex. The judge accepted that where the expression cannot be separated from the idea, there would be no protection, but found that in this case the words used in DataFlex go beyond that, and that each was an original expression and therefore subject to copyright. He specifically stated that because the Australian Act contains no equivalent to Section 102(b) of the US Copyright Act, analysis concerning the 'method of operation' (which is discussed in the American cases) could play no part in his decision. The decision is now suspect in light of Australia's acceptance of TRIPS, which has at Section, Article 9.2, that copyright is no extended to 'methods of operation'.

"The idea/expression dichotomy in the computer age" by Judge Jon Newman of the Court of Appeals of the Second Circuit (the Altai court), 17 Cardozo Arts & Ent. Law Journal 691.

Judge Newman notes the many problems of applying copyright law to software, for example, noting that computer programs and literary stories have very different structures, and also notes one of the criticisms of Altai. He suggests one way to solve the problem is to stop using "idea" and "expression" labels for aspects of software. Instead, he suggests to no longer use "expression", but rather "source code", and no longer use "idea" but rather "program purposes" or "program objective". He doesn't suggest using "program process result", because to do so forces him to mention what he doesn't mention anywhere in the paper, patents, maintaining the copyright world's fiction that the patent system doesn't exist. With 90,000+ software patents issued as of 1999, for all aspects of software from data structures upwards, to ignore how the useful processes and systems of these software patents severely limits any possible aspects of software to be protected with copyright is inexcusable. He ends the paper, seven years after his court created the abstraction-filtration-comparison test in Altai, by questioning the decision: "... and I am not sure we ought to have something called an 'abstraction-filtration-comparison test.". Why no mention of patent claims and Markman hearings?

Cantor Fitzgerald International v. Tradition (UK Times 19 May)

Extension of bogus Whelan logic into a UK decision. As part of the decision, the judge also accepted that the overall structure of a computer system or program could be protected in its own right if it is the result of sufficient skill, labor and judgment.

UCITA - Uniform Computer Information Transactions Act

An attempt by the software industry to update the Uniform Commercial Code to make software licenses more favorable to the software industry and much less favorable to consumers. The American Law Institute, which helped draft UCITA has withdrawn its support. The two main engineering and computer societies, the IEEE and the ACM (which have many lawyer members on their IP committees), oppose UCITA, as well as the Electronic Frontier Foundation, and the insurance industry (including the Tort and Insurance Practices section of the American Bar Association).

I will leave it to history to assess as irony that the last name of the (hopefully discredited) UCITA's principal author is the same as the last name of a leading author of the (should be) discredited CONTU report twenty one years earlier - Nimmer.

Bernstein v. US Department of State (9th Circ. - source code is speech)

Well critiqued in "When efforts to conceal may actually reveal". see {DIXON}. Also well critiqued in "Patenting speech" - "this attempt to equate functional texts with source code is fundamentally flawed because it conflates two different meanings of the word 'instruction'", that the First Amendment reaches expressive parts of functional works - see {BURK} In his paper, Burk cites a paper by Lemley and Volokh ("Freedom of speech and injunctions in IP cases", 1998 Duke Law Journal, 147) in which they argue that patent law should treat executable software as a virtual machine, while treating source code as copyright-protected expression. Not only is the distinction artificial, but easily circumventible (like using the First Amendment to share expressive source code), and makes a mess of the doctrine of equivalents.

Junger v. Daley (6th Circ. - source code is speech)

Ruled that software has some non-functional expressive aspects entitled to constitutional protection. Well critiqued with Bernstein in "Patenting speech" - see {BURK}.

"The evolving common law doctrine of copyright misuse: a unified theory and its application to software", by Brett Frischmann and Dan Moylan, Berk.Tech.L.J., Fall 2000

Dealing with patent misuse is a well-established doctrine applied by the courts. However, as the authors point out, copyright misuse is far from well-established, including for software. Instead of arguing that this is one more reason why software should not be protected by copyright, the authors (citing without criticism the messes caused by CONTU) show how to make misuse work better for software copyright.

"Guide to statutory protection for computer software in the US", law firm of Ladas & Perry,

An uncritical review of software copyright history and cases. Discusses 1980 Copyright Act, but none of its criticisms. Doesn't even mention CONTU, on which the 1980 Act rests. Discusses Whelan, Autoskill, and Altai without mentioning the many criticisms of both (see above). Discusses Altai without suggesting that it is a patent test using different words (abstraction being reverse engineering returning patentable data structures). Nice review of a variety of other aspects of software copyright implicitly assuming the legitimacy of software copyright. A good example of a law firm Web publication on software copyright that needs to reveal the many criticisms and weaknesses of software copyright.

18 Computer Scientists' Amici Curiae brief for DeCSS case

The section on software copyright approvingly cites all of the software copyright decisions that have been overturned and/or ignored such as Whelan, while ignoring the decisions contrary to software copyright. Since these omissions were deliberate, it questions the ethics and legal research skills of these 18 computer researchers.

"Code as speech: a discussion of Bernstein v. USDOJ, Karn v. USDOS, and Junger v. Daly in light of the U.S. Supreme Court's recent shift in Federalism", J. Camp and K. Lewis, Ethics and Information Technology, 2001, 21

The authors support the argument in these three cases that source code is speech, an unsupportable argument well critiqued in {BURK}. The lower court in Junger had it right - source code is action that does not rise to the level of speech (thereby falling under the ruling of O'Brien [391 US 367] and not being accorded the full protection of the First Amendment). The authors disagreed with one judge's view that 'source code and object code are interchangeable', saying "Judge Gwin's assertion ... is simply wrong". No, Gwin is mostly right, especially if the authors had given any consideration (ignored in the article) to the process of reverse engineering (which can take assembly and object code and abstract upwards to system specifications - there is an IEEE conference just on this topic which the authors, members of the IEEE, should have paid attention to). They have other assertions about software also suspect, weakening their arguments supporting the software as speech position.

In general:

Today's software copyright (US/NAFTA/TRIPS) is based on the often criticized vague 1980 Copyright Act amendment, which adopted the discredited 1978 CONTU report, which refused to challenge the 1972 Gottschalk/Benson anti-software patent ruling, an extremely weak legal ruling relying mostly on the anti-software patent wishes of large computer hardware manufacturers as biased into a 1966 Presidential commission report. In short, today's global software copyright laws are not based on logic or science, but the forty year-old anti-patent wishes of hardware companies who now increasingly earn their profits from their non-hardware products and services.

Software copyright, where it is not illogical, is based on outdated economic policy considerations. Globally, software copyright laws should be abolished.



Stefan Bauer-Mengelberg
"Parker v. Flook: a formula to cause alarm"
IDEA, 19XX, 349 (anyone with a formal cite for this, email me)


Dan Burk
"Patenting speech"
Texas Law Review, November 2000, 99


Donald Chisum
"The Patentability of Algorithms"
University of Pittsburgh Law Review, 1986 v47, 959


Anthony Clapes and Jennifer Daniels
"Revenge of the Luddites: a closer look at Computer Associates v. Altai"
Computer Lawyer, November 1992, 11


Rod Dixon
"When efforts to conceal may actually reveal: whether First Amendment protection of encryption source code and the Open Source movement support re-drawing the constitutional line between the First Amendment and copyright"
Columbia Science and Technology Law Review, Volume I, 2000


Dunner, Gambrell, White & Kayton
"Nonstatutory subject matter"
Jurimetrics Journal, 1973, 113


James Goodman, Todd Marlette, Peter Trzyna
"Toward a fact-based standard for determining whether programmed computers are patentable subject matter: the scientific wisdom of Alappat and ignorance of Trovato"
J. Patent Trademark Office Society, May 1995, 353


Robert Gorma
"Commentary: Comments on a manifesto concerning the legal protection of computer programs"
Albany Law Journal of Science & Technology, 1996, 277


Mark Haynes
"Black holes of innovation in the software arts"
Berkeley Technology Law Journal, Spring 1999, 503


Cary Kappel
"Copyright protection of SSO: replete with internal deficiencies and practical dangers"
Fordham Law Review, 1991, 699


Dennis Karjala
"Copyright protection of computer documents, reverse engineering and Professor Miller"
Dayton Law Review, 1994, 975


Dennis Karjala
"A coherent theory for the copyright protection of computer software and recent judicial interpretations"
University of Cincinnati Law Review, 1997 v66, 53 [Despite title, concludes most of software is noncopyrightable, except maybe for user interfaces, which was later denied in Borland V. Points out an incompetence of the copyrighters.]


Robert Laurenson
"Computer software 'article of manufacture' patents"
J. Patent Trademark Office Society, Oct 1995, 811


"Software Protection Integrating Patent, Copyright and Trade Secret Law"
J. Patent & Trademark Office Society 1987, 151


Leo Raskind
"The uncertain case for special legislation protecting computer software" University of Pittsburgh Law Review, 1986 v47, 1131
Argues against sui generis protection for software, but still finds some aspects of software worth protecting.


John Rauch
"The realities of our times: the Semiconductor Chip Protection Act of 1984 and the evolution of the semiconductor industry"
Journal Patent and Trademark Office Society, February 1993, 93


Michael Sutton
"Equities, evidence and the elusive scope of copyright protection for computer software"
Journal of The Patent Trademark Office Society, October 1987, 551


Robert Shaposka
"Note: A criticism of the E.U. directive protecting computer software"
Buffalo Journal of International Law, Winter 1996-7, 519

The expanded version of this will have one other section - a review of the physics and mathematics of software as purely patentable process and machine.


Synercom Technology v. University Computing (N.D. Tex. - 462 F. Supp. 1003)

District court held that "the sequence and ordering" of input data was idea rather than expression and therefore not copyrightable.

System Science v. Toyo Sokki (Japanese courts)

Processing flows of a subroutine, and its structure, is an algorithm beyond copyright protection.

U.S. Copyright Office (54 Fed.Reg. 13179 [1989])

The US Copyright Office refuses to register for copyright elements of software as allowed by Whelan: "The Office will not make registration based solely on a claim of structure, sequence and organization. These terms are ambiguous, and may refer to uncopyrightable ideas, systems, or processes. Thus, claims described as so-called SSO are unacceptable."

Telemarketing Resources v. Symantec Corp. (DCNDCA, 12 USPQ.2D 1991 [1989])

Popular aspects of user interfaces (menu screens, menu bars, pull-down windows, color schemes) are not copyrightable.

NEC v. Intel (10 USPQ 2d)

Microcode for the instruction set of a microprocessor is not protected by copyright (though sadly relying on the merger doctrine).

Total Information Processing Systems v. Daman (UK FSR 171)

UK judge rules that the structure of a database (the fields and records specifications) were not copyrightable. He also rules that a compilation of several computer programs is also not copyrightable. The ruling, though, is not based on functionality or idea considerations, but amount of labor efforts.

Gates Rubber v. Bando Chemical Industries (10th Circ. - 9 F.3d 823)

A decision with some problems (see above), it did rule that:
- main purpose or function of a computer program is not copyrightable
- purpose or function of a computer module nearly always not copyrightable
- constants in formulas are facts not copyrightable

Apple Computer v. Microsoft and HP (9th Circ. - 35 F.3d 1435)

Arrangement of unprotectible elements of user interface design are not copyrightable, short of virtual identity. Virtual identity is always easy to workaround, leaving nothing to protect with copyright.

IGM v. Mets (Japanese courts)

MSDOS batch file commands were not sufficiently creative to be expression, but rather a data file that did not qualify for copyright protection.

Lotus v. Borland (1st. Circ. - 49 F.3d 807)

Software menu structures are "means of operation" that are not copyrightable. "The 'expressive' choices of what to name the command terms and how to arrange them do not magically change the uncopyrightable menu command hierarchy into copyrightable subject matter."

Baystate Technologies v. Bentley Systems (D.Mass - 946 F.Supp. 1079)

Data structures, their names and organization of those names (in this case, data file elements) are not copyrightable. Which makes sense, since in 1994 the CAFC ruled that data structures are patentable (In re Lowry, 32 F.3d 1579), and therefore useful articles. Baystate came after the First Circuit opinion in Lotus v. Borland that menus are not copyrightable.
"Comprehensive analysis of software "look and feel" protection" by David Hayes of Fenwick and West
Argues what little of software "look and feel" is protectable by copyright is analyzed by the courts using Altai. But given that Altai is a bad attempt to make copyright law work by using patent principles, one wonders if anything part of "look and feel" is, or should allowed to be, protectable by copyright.

Computer Management Associates v. Robert DeCastro Inc (5th Circ. - 220 F.3d 396)

Court rules that "scenes a faire" doctrine excludes from copyright protection work serving functional purposes or work that is dictated by external factors such as particular business practices; in area of computer programs, these external factors may include hardware standards and mechanical specifications, software standards and compatibility requirements, computer manufacturer design standards, target industry practices and demands, and computer industry programming practices.