en.swpat.org is a wiki.   You can edit it.   May contain statements End Software Patents does not endorse.

October 2014: US Supreme Court reining in software patents (Alice v. CLS)!

SitemapCountriesWhy abolish?Law proposalsStudiesCase lawPatent office case lawLawsuits


Shielding software from litigation

From en.swpat.org
Revision as of 13:03, 14 November 2012 by Gibus (Talk | contribs)

Jump to: navigation, search

Another approach to ending the problems of software patents would be a law saying, as Richard Stallman puts it, "that developing, distributing, or running a program on generally used computing hardware does not constitute patent infringement."[1]

(The other approach is excluding software from patentability.)

Contents

Advantages

  1. The main advantage is that this avoids the need for a legal definition of "software patent". Patents no longer have to be classified as software or non-software. "Software" patents can still be granted, but no patents can be used to attack people for developing or distributing software.
  2. Interpretation of the law lies directly with the judge, instead of the patent office.
  3. The wording of the patent is no longer important, so the ingenuity of patent lawyers for finding loopholes and confusing wordings can no longer influence the chances of a lawsuit succeeding against a software developer.
  4. This creates a clear situation for already-granted software patents, rather than leaving a question mark over whether they would be upheld or invalidated by a court after a change in patent eligibility criteria.

For the EU unitary patent

For the EU regulation on the unitary patent, an amendment to exclude software from patentability has already been proposed (see amendments 5 and 6). It has been objected that this regulation was not about (re-)defining rules on patentability. But in this regulation there is already an article (Art. 8) defining exceptions to patentees' rights: for non-commercial private acts; experimental uses; medicines during patent term to obtain regulatory approval; extemporaneous preparations in a pharmacy of medicines; acts on or concerning foreign means of transport which temporarily or accidentally enter EU; acts as covered by the farmers privilege; some defined biotechnological acts; and decompilation of software in order to obtain informations necessary for interoperability.

An amendment implementing Stallman's suggestion for an exception for "developing, distributing, or running a program on generally used computing hardware" would be easier to be adopted.

What is "generally used computing hardware"?

Richard Stallman says that this exception would not apply to software on "special-purpose hardware". I wonder how this affects the overall outcome... but Stallman doesn't see this as a problem, so maybe it isn't:

This approach doesn’t entirely invalidate existing computational idea patents, because they would continue to apply to implementations using special-purpose hardware. This is an advantage because it eliminates an argument against the legal validity of the plan.

Let's take the example of software patents on MP3.

If Stallman's proposition to limit effects of software patents only to "generally used computing hardware" is passed, sellers of MP3 reading devices (hardware) will still have to pay a license (the price being transferred fully or in part to consumers) to MP3 patents' holders. The latter could still file patents on compression/decoding of audio formats. The MP3 algorithms are implemented on circuits, that is on "special-purpose computers". The exception does not apply.

Conversely, developers and users of software for reading MP3 would not need to fear the existing patents, nor future patents. Such software is implemented on "generally used computing hardware", that is "general-purpose computers". Developing, distributing and running such software on general-purpose computers is covered by the exception.

This effect is very different from a solution based on the exclusion of software from patentability. If such an exclusion is passed, nobody could obtain a patent on encoding/decoding audio algorithms anymore (e.g. on an imaginary MP138 format). Makers of MP138 reading devices would not have to worry (and consumers wouldn't see any change); nor would developers and users of software for reading MP138.

Conversely, holders of existing software patents on MP3 might want to continue to get some royalties, both from makers of MP3 reading devices (with cost transfered to consumers), and from developers and users of software for reading MP3. If there is litigation, these patents might be invalidated, since they had been mistakenly granted on excluded subject matter.

Why a limited exception?

One could ask why limiting the exception to software intended to run on a "generally used computing hardware", or a "general-purpose computer", and not having this exception for all software, even if implemented on a special-purpose computer, on a circuit, etc.

The main reason, as stated by Stallman in his column on WiReD is that:

This is an advantage because it eliminates an argument against the legal validity of the plan.

Indeed, one recurring argument against the solution to exclude software from patentability is that other non-software patents could be affected altogether with such an exclusion. For instance, we don't really care if patents on a rubber-curing machinery, like the one at stake in Diamond v. Diehr, can still be enforced on computational ideas implemented on a rubber-curing machinery. We just want to be free to implement, distribute and use the same computational idea in software on our usual computers.

Another reason can be see in Article 30 TRIPS:

Members may provide limited exceptions to the exclusive rights conferred by a patent, provided that such exceptions do not unreasonably conflict with a normal exploitation of the patent and do not unreasonably prejudice the legitimate interests of the patent owner, taking account of the legitimate interests of third parties.

Limiting the exception to acts on general-purpose computers can help to comply with requirements of Article 30 TRIPS. Indeed the first requirement that the exception should be limited would not be met if the exception was too broad to encompass any acts related to computational ideas. About the other requirements, it is not certain whether they are met or not. But with political will, the whole test of Article 30 is very likely to be passed. See about this the study by Professor Lionel Bently for WIPO Standing Committee on the Law of Patents: Exclusions from Patentability and Exceptions and Limitations to Patentees’ Rights:

In addition, it should be acknowledged that our experience with exceptions is rather limited. One particular concern is precisely how much flexibility is left to national authorities as a result of Article 30 of TRIPs (in combination with the non-discrimination provisions of Article 27). Although that provision is somewhat open-textured, further guidance as to the level of flexibility it offers could be of real value. The WTO Panel in Canadian Pharmaceuticals gave helpful findings on the notion of “discrimination” between fields of technology, but other aspects of the holding (for example, that on the notion of “limited”) may operate to deprive member countries of the real potential offered by the use of exceptions. In our view, this would be regrettable.

Choice of wording

As seen above (with examples of "generally used computing hardware" or "general-purpose computer", if we propose some legislation for this exception, we can choose between different wording to limit the exception to some specific acts. What is important is that everything that we want to be shielded from software patents enforcement is covered by the exception, while we can live with everything not falling in the exception.

In WiReD, Stallman has chosen the wording "generally used computing hardware". From some personal exchange with him, his main intention is:

The idea of "generally used computing hardware" is to distinguish them from cases like the rubber-curing plant in Diamond v. Diehr. The precise words to use here call for some thought. We need to think about which cases should be covered by the exception, and how to express that line clearly.

Gibus who is working on https://www.unitary-patent.eu has proposed to use the wording "general-purpose computer".

First this expression helps indeed to distinguished covered acts from those on a rubber-curing machinery in [[Diamond v. Diehr, precisely because U.S. Supreme Court has said it was patentable, while refusing patentability in Gottschalk v. Benson, using in that case the expression "general-purpose digital computer". While the only place where "general-purpose digital computer" is used in Diamond v. Diehr is to reiterate the rejection of patentability of Gottschalk v. Benson.

Second because any wording that includes "hardware" would prevent the exception to apply on acts performed on virtual machines. For instance, using Kernel-based Virtual Machine (which is available in the Linux kernel), one can emulate several virtual machines, actually running on the same physical machine. Different distributions of GNU/Linux can be run simultaneously on different virtual machines. With "generally used computing hardware", running GNU/Linux or any software on a virtual machine would not be covered by the exception. Conversely, using "general-purpose computer" allows to cover software run on a virtual machine, which is indeed a general-purpose computer implemented in software.

Third, the definition of "general-purpose computer" is non ambiguous and well-known, both in technical and legal literature. According to the EU inter-institutional terminology database, the definition from ISO/DIS/I is

a computer that is designed to operate upon a wide variety of problems

The exact same definition seems also to be found on American National Dictionary for Information Processing, by the Canadian Government EDP Standards Committee), or by Computer Science and Communications Dictionary, by Martin Weik. Very close definitions can be found in other various dictionnaries.

Moreover "general-purpose computer" is widely used in legal documents, notably on many patents and patent applications in Europe and US, on some memorandums of the European Commision, or rulings of the US Supreme Court or of the Court of Justice of the European Union, and even by Richard Stallman (stating that “Treacherous computing” is designed to stop your computer from functioning as a general-purpose computer).

Richard Stallman has objected to the second argument above that actually with "generally used computing hardware", running GNU/Linux or any software on a virtual machine would not be covered by the exception, because:

The specific program, and the emulator, are a collection of software and it is running on generally used computing hardware.

It may be perfectly exact. Nevertheless, the mere fact that there could be a doubt tends to show that this is a wording which can be disputed in court.

Stallman has then proposed the following wordings:

However, perhaps "non-task-specific computer hardware" is better than "generally used computer hardware". Or "not specific to the purpose achieved by the patented technique".

"non-task-specific computer hardware" would be equivalent to "not special-purpose computer", i.e. "general-purpose computer".

It would actually be dangerous to use the last proposition in a legal context in Europe, as the word "technique" would validate that software is a technical field, and therefore would block the solution of excluding software from patentability.

Such wording could nevertheless be used for communication to public.

There is a concern that expressions like "general-purpose computer" or whatever wording equivalent to a Turing machine would be interpreted as non including ebook readers, routers, camera, circuit implementing MP3 decoding algorithm, etc.

In a legal context, the last proposition above could actually be used by replacing "technique" by the legally defined term "subject matter". This is too legalistic and clunky for talking to the public, but accurate for a legal proposition.

So the best wording found until now for a legal proposition would be "computing equipment that is not specialized for implementing the patented subject matter".

However, we need to have a lot of people to look at each proposal, to look for what its consequences would be in various cases, so we can think about which criterion is right.

Current shields for other domains

The legality of this approach is confirmed by the existence of similar protections for other domains.

US surgeon's precedent

Such a shield already exists since 1997 for surgeons in the USA in 35 U.S.C. Section 287 (c), sometimes called the "Ganske compromise legislation":

(c)
(1) With respect to a medical practitioner’s performance of a medical activity that constitutes an infringement under section 271 (a) or (b), the provisions of sections 281, 283, 284, and 285 shall not apply against the medical practitioner or against a related health care entity with respect to such medical activity.
(2) For the purposes of this subsection:
(A) the term “medical activity” means the performance of a medical or surgical procedure on a body, but shall not include
(i) the use of a patented machine, manufacture, or composition of matter in violation of such patent,
(ii) the practice of a patented use of a composition of matter in violation of such patent, or
(iii) the practice of a process in violation of a biotechnology patent.
(B) the term “medical practitioner” means any natural person who is licensed by a State to provide the medical activity described in subsection (c)(1) or who is acting under the direction of such person in the performance of the medical activity.
(C) the term “related health care entity” shall mean an entity with which a medical practitioner has a professional affiliation under which the medical practitioner performs the medical activity, including but not limited to a nursing home, hospital, university, medical school, health maintenance organization, group medical practice, or a medical clinic.
(D) the term “professional affiliation” shall mean staff privileges, medical staff membership, employment or contractual relationship, partnership or ownership interest, academic appointment, or other affiliation under which a medical practitioner provides the medical activity on behalf of, or in association with, the health care entity.
(E) the term “body” shall mean a human body, organ or cadaver, or a nonhuman animal used in medical research or instruction directly relating to the treatment of humans.
(F) the term “patented use of a composition of matter” does not include a claim for a method of performing a medical or surgical procedure on a body that recites the use of a composition of matter where the use of that composition of matter does not directly contribute to achievement of the objective of the claimed method.
(G) the term “State” shall mean any State or territory of the United States, the District of Columbia, and the Commonwealth of Puerto Rico.
(3) This subsection does not apply to the activities of any person, or employee or agent of such person (regardless of whether such person is a tax exempt organization under section 501(c) of the Internal Revenue Code), who is engaged in the commercial development, manufacture, sale, importation, or distribution of a machine, manufacture, or composition of matter or the provision of pharmacy or clinical laboratory services (other than clinical laboratory services provided in a physician’s office), where such activities are:
(A) directly related to the commercial development, manufacture, sale, importation, or distribution of a machine, manufacture, or composition of matter or the provision of pharmacy or clinical laboratory services (other than clinical laboratory services provided in a physician’s office), and
(B) regulated under the Federal Food, Drug, and Cosmetic Act, the Public Health Service Act, or the Clinical Laboratories Improvement Act.
(4) This subsection shall not apply to any patent issued based on an application the earliest effective filing date of which is prior to September 30, 1996.

Related pages on en.swpat.org

External links

References

  1. http://www.wired.com/opinion/2012/11/richard-stallman-software-patents


This wiki is part of the End Software Patents (ESP) campaign (donate). For more information, see:
>> endsoftwarepatents.org (Main ESP website) <<
>> endsoftwarepatents.org/news (News) <<

This wiki is publicly editable. (See: en.swpat.org:About) It's a pool of information, not a statement of ESP's views or policies, so no permission is required. Add your knowledge! (See: Help:How to make a good contribution)