copyright 1995 Stephen C. Glazier
Keck, Mahin & Cate ; Washington, D.C., U.S.A.
This article is a chapter from Mr. Glazier's new book entitled "Patent Strategies for Business" published by Euromoney Books. For further information on how to order the book.
The first ten months of 1995 brought several important legal developments to U.S. intellectual property law. These changes represent the continuing strengthening of software patents in the United States (and perhaps a corresponding weakening of software copyright), and the further development of U.S. patent law to harmonize with trends in the rest of the world.
On June 27, 1995, the U.S. Patent Office published in the Official Gazette proposed "Examination Guidelines for Computer Implemented Inventions" (the so-called "Alappat Guidelines"). These are proposed guidelines to be issued to Patent Office examiners for the review of software oriented patent applications. These guidelines were issued in response to the capitulation of the Patent Office to the Federal Circuit and the pro-software patent trends manifested In Re Alappat (discussed elsewhere herein).
These guidelines represent a major new pro-patent step forward by the Patent Office. These guidelines will go a long way to resolve the confusion at the Patent Office and in the patent profession as to the best way to format claims for software applications. This in turn will accelerate, standardize, and make easier for applicants, the entire software patent application procedure. In the end, this will make software patents more common, and more valuable.
Perhaps the biggest step forward in the Alappat guidelines is implicit. The guidelines can be taken basically as an instruction from the Commissioner to the examiners to cooperate and facilitate in the examination of software patents.
The Alappat guidelines explicitly direct examiners to deal with software patent applications (so-called "computer implemented inventions") in a "prompt yet complete examination". The guidelines also direct examiners to not limit themselves to just format objections to claims in the initial office action, but also to deal with substantive patentability issues. This alleviates a problem of a lot of software prosecution lately at the U.S. Patent Office, where initial office actions concern themselves mainly with the format of the software claims, and do not deal with the substance of patentability. These "format only" actions led to slow prosecution of software applications. Software patent prosecution was additionally confused because of the lack of agreement among examiners, and lack of agreement between the Patent Office and the Federal Circuit, over the proper form of software claims. Now, under these guidelines, examiners would be directed to bring up format issues in the indicated uniform manner in their first action, and also to deal with substantive patentability problems.
Furthermore, the guidelines direct examiners to "indicate how rejections may be overcome and problems resolved", so that patents may be issued more readily. That is, examiners should indicate where possible, in each official response, how to proceed to obtain a patent. The role of the Patent Office is, then, not to say "no", but to help get to "yes".
The guidelines also go a long way to clarify the proper possible formats for software claims. To be in proper format, a claim must on its face describe a technology that is eligible for a patent under the patent statute. For example, a "statutory machine" is eligible, and a claim in that format may be considered. Since "software" is not mentioned in the statute, the game for software claims must be to format the claims as one of the statutory technologies. Three basic formats are presented:
1. Software Machines . The guidelines indicate that "a computer or other programmable apparatus whose actions are directed by a computer program or other form of software is a statutory machine." Hence, a software claim may be in proper format if it reads something like:
a programmable apparatus for [doing a function] comprising:
(1) means for [doing a function element]
(2) etc.
Apparently, this would also be adequately claimed with claims in the format of:
a programmable device whose actions are directed by software executing a process comprising:
(1) doing [step 1]
(2) etc.
This latter formulation might also avoid the sort of Donaldson" problems of interpreting "means plus function" claims, described later herein.
2. Software Processes . A process format for software claims may also work; that is, a "series of specific operational steps to be performed on or with the aid of a computer is a statutory process". Presumably a claim in the following format would fall in this category:
A process to be performed on or with the aid of a computer, comprising the following steps:
(1) doing [step 1]
(2) etc.
These guidelines clarify classic arguments over process and apparatus formats for claims for software inventions. In the past a lot of confusion has arisen regarding software claims formatted as apparatus claims and/or as process claims. Now these guidelines tell us that both formats work well, and instructs us in how each format should be structured.
The preceding developments in the guidelines are an excellent contribution to the field and will do a lot to expedite and make more consistent software applications within the Patent Office.
3. A Great New Format: Software in Memory:
A third format presented in the guidelines was somewhat of a surprise. The new format is a great innovation and will be a tremendous help to the profession. The third format offered by the guidelines states, "A computer readable memory [containing software] that can be used to direct a computer to function in a particular manner when used by the computer is a statutory article of manufacture". The guidelines elsewhere go on to state that "a computer readable memory encoded with data representing a computer program that can cause a computer to function in a particular manner or a component of a computer that can be reconfigured with a computer program to operate in a particular fashion can serve as the specific structure corresponding to a means element". That is, "computer program related elements of a computer implemented invention may serve as the specific structure, material or act that correspond to an element of an invention defined by a means plus function limitation".
Therefore, the following claim format should be a statutory article of manufacture:
"A computer readable memory, encoded with data representing a computer program, that can be used to direct a computer when used by the computer, comprising:
(1) Means for [a function]
(2) etc.
The guidelines would make software claims in this new format much easier to enforce. In the case where patentable software is pirated and sold by a middle man pirate, to legitimate customers of the patent holder, the pirate himself may not be a direct infringer of software claims in the classic formats (that is, claims that are limited to a computer programmed with the software, or the process as executed on a computer by the software). That is, the pirate may have neither the hardware platform to execute the software nor ever actually execute the software, but merely have the equipment to copy the software and the copies of the software on discs. Hence, to pursue an infringement claim against the true culprit, the pirate, a patent holder would have to presumably sue its ultimate legitimate client, the buyer, for infringement and go against the pirate for contributory infringement or induced infringement.
However in the new format, with claims for the software on a disc, the patent holder could sue the middle man pirate only as a direct infringer.
The guidelines go on to indicate that data structure can be patented and how this may be done.
The guidelines would simply state that "a claim that defines an invention as any of the following subject matter should be classified as non-statutory [that is, unpatentable] . . . [for example] a data structure independent of any physical element (i.e., not as implemented on a physical component of a computer such as a computer readable memory to render that component capable of causing a computer to operate in a particular manner. . .)"
Therefore, the direction is that a data structure with any physical element (for example, as implemented on a physical component of a computer such as a computer readable memory to render that component capable of causing a computer to operate in a particular manner) would be statutory, that is, patentable. Hence, a data structure claim of the following format should be a patentable article of manufacture, assuming that it meets the other requirements of being new, useful, and not obvious:
a computer readable memory containing data with a structure capable of causing a computer to operate in a particular manner, the structure comprising:
Element 1.
Element 2.
etc.
Other comments in the Patent Office legal analysis of the Alappat guidelines, published 3 October 1995, indicate that computer data structure can be claimed by the process of structuring that data in a computer. Also, it is indicated that particular data loaded into a computer for a particular function can also constitute a patentable machine.
The guidelines further state that "an applicant may not use computer program code, in either source or object format, to define the meets and bounds of a claim". Of course, source code or flowcharts or object diagrams or other disclosure could be used, and may be required in some instances, in the specification as support for the claims.
It is further interesting to note, the comment and the guidelines that: "articles of manufacture accomplished by this definition [regarding computer readable memory] consist of two elements:
1. A computer readable storage medium such as a memory device, a compact disc or a floppy disk, and
2. The specific physical configuration of the substrate of the computer readable storage medium that represents data (for example, a computer program) where the storage medium so configured causes a computer to operate in a specific and pre-determined manner."
The composite of the two elements is "a storage medium with a particular physical structure and function (for example, one that will impart the functionality represented by the data onto the computer)."
Note that the suggestion here is that data imparts functionality as much as does source code, so that data on a disk should be as patentable as source code on a disk.
Furthermore, note the interesting drafting tip that "for example, a claim that is cast as a 'computer program' but which then recites specific steps to be implemented on or using a computer, should be classified as a 'process'. A claim to simply a 'computer program' that does not define the invention in terms of specific patent steps to be performed on or using a computer should not be classified as a statutory process."
Also note the interesting comment that "the specific words or symbols that constitute a computer program represent the expression of the computer program and as such are a literary creation". Hence, they would not be protectable by patent, but would be protectable by copyright, to the extent the copyright protects at all.
It is further interesting to note that the In Re Alappat guidelines do nothing to stop the slow death of the business method doctrine. The guidelines state "non-statutory subject matter (that is, abstract ideas, laws of nature and natural phenomena" . . . note that this list does not include "business methods". The implication is that the Patent Office is taking no position to support the idea that business methods are non-statutory subject matter. This is important because this common law idea is an attack used against software patents. As suggested by the dissent in In Re Schrader (discussed elsewhere herein), the Federal Circuit may eventually directly eliminate this doctrine as an attack on software patents, and on any other type of patent.
In many applications, it would be anticipated that claims for basically the same invention would be each stated in all three formats. Of those three clone claims, the software in memory might be the broadest and most valuable.
Of course, all these claim formats would require adequate support and description in the specification, regarding enablement and best mode, and they would have to be new, useful and not obvious. These remain requirements of all patents.
It is also quite interesting that the guidelines further state as directions to examiners to "construe any element defined in means plus function language to encompass all reasonable equivalence to the specific structure, material or acts disclosed in a specification corresponding to that means element". That is, means plus functions claims are to be construed by the Patent Office so as to be limited by the specific disclosure in the specification, and are not to be construed in the much broader fashion to include any possible means executing that function.
This is consistent with 35 U.S.C. 112(6), but following this policy is a new reversal of prior policy at the Patent Office. Previously, the Patent Office had caused considerable confusion by insisting on interpreting claims for examination purposes in means plus function format to include all possible means, and not to be limited by the specification as required by statute. This would make claims appear much broader than necessary and has made it quite difficult for certain software claims to be allowed. This policy had been pushed by the Patent Office contrary to clear case holdings by the Federal Circuit. This was a fundamental policy split between the Patent Office and the Federal Circuit and caused considerable confusion for software patents.
However, this explicit Patent Office policy to follow the statute and the Federal Circuit cases on this point, will make it much easier to get means plus function format software claims allowed, if they are properly written and supportive specifications are included.
This policy in the guidelines is a reference to the policy adopted in more detail in the so called In Re Donaldson examination guidelines, published May 17, 1994, in the Official Gazette of the U.S. Patent Office. These guidelines were written in response to In Re Donaldson, 29 U.S.P.Q. 2nd 1845 (Fed. Cir. 1994).
The In Re Alappat guidelines together with the In Re Donaldson guidelines, represent a tremendous step forward by the Patent Office to adopt a pro-software patent stance and to end its struggle against the earlier leadership of the Federal Circuit in this regard. This final pro-patent consensus between the U.S. Patent Office and the Federal Circuit establishes a uniform national policy in the United States that is pro-software patent and represents a considerable legal and institutional clarification and unification of the law. With the In Re Alappat guidelines, it is now clear that the split between the Patent Office and the Federal Circuit on this point is over and practice in the area should be tremendously facilitated.
On October 3, 1995, the Patent Office issued a legal analysis supporting the proposed examination guidelines for computer implemented inventions. This was an effort by the Patent Office to provide extensive legal support for its proposed guidelines.
Another important development in 1995 was the effectiveness of the GATT amendments to the U.S. Patent Statute. These took effect for applications filed after June 8, 1995.
Perhaps the most interesting GATT amendment to the patent statute is the change of the term of a U.S. patent from 17 years from the date of issuance, to 20 years from the date of application. Foreign priority under Section 119 is not taken into account in this calculation of term.
Since the date of expiration of an eventual patent is now fixed when the application is made, procedural steps that may expedite issuance of patents may now be more popular. In particular, the traditionally obscure procedure for expedited review of patent applications, the so-called Petition to Make Special, may now come into more common use. This is because obtaining earlier issuance of a patent will extend the period of enforceability. Likewise, divisional applications may be more common, as early allowed claims are taken and rejected claims are continued to be argued for.
The shorter patent term may cause "submarine" patent applications, in extreme examples, to surface more quickly, and certainly before they entirely expire. However, the only impact on the submarine patent strategy will be that the submarine can not stay submerged for more than 20 years.
The amendments do not cause publication of pending U.S. patents, or provide for opposition proceedings against pending U.S. applications by third parties. However, these policies are proposed and supported by the Commissioner of Patents. Hopefully, these changes will come about in the United States, but it is not possible to accurately predict when this will happen. Publication of pending applications and permitting third party oppositions (perhaps through expanded re-examinations) would be especially useful in getting more valid and stronger software patents. This is because it is so difficult for the Patent Office or anyone else to do effective prior art searches in the software area. Hence, software patent applicants can look forward to the opportunity of publication and opposition to U.S. software patents, and if and when this happens, U.S. patent practitioners with experience in European and Japanese opposition proceedings will be of particular interest in the United States.
The new GATT amendments also allow for provisional applications. The new section 35 U.S.C. 111(b) allows for provisional applications that meet the requirements for an application under section 112, but lack claims. A complete section 112 application filed within a year claiming the priority date to the provisional application will be prosecuted with that priority date. This may become a handy procedure in certain circumstances when a publication or on-sale bar may not be avoidable, but a formal complete patent application is not ready for filing.
Regarding foreign inventions, a patentee may now provide a date of invention by acknowledged use or other activities in any WTO ("World Trade Organization") member country (over 100 countries have now become WTO member countries). Before this GATT amendment, a date of invention could only be based on U.S. activities, or since 1994 and the NAFTA agreement, activities in Canada and Mexico. This allows inventors applying in the United States to prove first invention by activities outside of the United States.
Regarding infringement, section 271 has expanded the definition of infringement. The old definition provided for that infringement was constituted by making, using or selling the claimed invention in the United States. Now, under the new amendments, offering for sale or importing is included as an infringing activity.
Also, note that other recent amendments provide that if process claims are in a patent, and a product is made by the claimed process but the product is not itself claimed, then using, selling, offering for sale, or importing into the United States an unpatented product made by the patented process will, in itself, constitute infringement. Although this was perhaps written contemplating biogenetic engineering inventions, it may have important implications in some cases involving software claims that assume a process format.
It is also interesting to note that the Patent Office is undertaking a lobbying and educational program with the European Patent Office ("EPO") and the Japanese Patent Office ("JPO"). The intent of this effort is to get the EPO and the JPO to adopt guidelines regarding software applications that are similar or the same as the Alappat guidelines. This would provide for a uniform and faster review of software applications in the three major patent offices in the world.
After In Re Alappat, 33 F.3d 1526 (Fed. Cir. 1994) was handed down, in which it became clear that software running on a general purpose computer was patentable subject matter, several cases followed. In Re Warmerdam, 33 F.3d 1354 (Fed. Cir. 1994), was decided by a three judge panel and was basically consistent with Alappat. Then In Re Lowery, 32 F.3d 1579 (Fed. Cir. 1994) was also decided by a three judge panel in a manner consistent with Alappat regarding software patents. However, it is almost amusing to note a later case, In Re Trovato, 42 F.3d 1376 (Fed. Cir. 1994), which was decided by a three judge panel, in which the majority on the panel were in the minority in Alappat. The decision in Trovato in effect attempted to overrule Alappat and state that software was basically not patentable, although many points of the decision were blatantly wrong on the facts. This case is probably bad law and may be reversed in an en banc rehearing. It is also interesting to note that In Re Beauregard, 53 F.3d 1583 (Fed. Cir. 1994), involved a software patent application which was rejected by the Patent Office. The patent applicant appealed. However, prior to a decision on the case, the Patent Office reversed its policy on software patents, issued the proposed In Re Alappat guidelines, and effectively dropped its opposition to the Beauregard application.
This trend represents the triumph of the software patent, and the resolution of the conflict between the Federal Circuit and the Patent Office in this regard. The Federal Circuit did a tremendous service in properly interpreting U.S. patent statutes regarding the new software technology and in leading the development of the case law in this matter. The Patent Office and particularly the new Patent Commissioner did a great service in finally adopting the wisdom of the majority of the Federal Circuit on this issue. With the new apparent consensus between the Patent Office and the Federal Circuit on the patentability of software, the last major conflict within the development of U.S. patent law regarding software patents has been resolved. We can now see entering a permanent Golden Age of software patents in the United States, typified by faster and cheaper and more predictable application and enforcement of software patents.
The final step to improve the situation would be to amend the patent statute to provide for the publication of pending patent applications and third party opposition to applications. This would solve the problem of the objective difficulty in doing prior art patentability searches for software. This would lend added teeth to the statutory presumption in the United States that issued U.S. patents are valid and, in the end, this would help the industry. The current U.S. Patent Examiner encourages publishing U.S. patent applications and in providing some sort of third party opposition procedure, but it remains to be seen if this change will be made. This would require and amendment to the U.S. patent statute by Congress. The proposals for these amendments are before Congress and we will have to watch developments to see if they are passed.
In this light, it is interesting to note the apparent further shrinking of copyright law to protect software.
A trend has developed in most of the Federal Circuits that copyright could protect the so-called look and feel of software, or the command structure and hierarchy of software. However, in 1995, in the case Lotus v. Borland, 49 F.3d 807 (First Cir. 1995), it was held that the Lotus 1,2,3 command structure could not be protected by U.S. copyright law because it was not copyrightable subject matter. The U.S. Supreme Court recently decided to hear the appeal of this case. Hopefully, the eventual U.S. Supreme Court decision on Lotus v. Borland will resolve the question of the copyrightability of software, particular its look and feel and command structure, and resolve the split in the various circuit courts in this regard.
This author feels that contrary to the majority of the circuits, the First Circuit is correct. Copyrights should correctly be applied to actual copying of source code, object code, and images generated and presented on computer monitors or other medium by the software, or other visual works generated and presented by the software. Conventional copyright analysis does not require to be changed to continue this protection. However, this author feels that the recent stretching of copyright law to protect the look and feel of software and command structure was a well meaning but erroneous attempt to stretch copyright law to make it more patent-like in its protection of software.
This expansion of copyright by judges took place before the question of the patentability of software was resolved by the courts and the Patent Office. However, ultimately this expansion of copyright should be seen as incorrect and, indeed, unnecessary in light of the clarification of U.S. patent law regarding software.
Patent protection of software is more difficult to obtain than copyright protection, but patent protection is more profound and generalized, and is independent of specific source code, object code, or visual works generated and presented by the software. Copyright protection is easier and more automatic to obtain, but is less effective and easier to go around.