Sharon Bowles Letter to Constituent: Comments
--> [ Toine Manders | UK Libdems | Sharon Bowles MEP | UK Parties | UK Patent Family Software Patents ]
The following is an annotated letter from UK Liberal Democrat spokesperson Sharon Bowles on software patents, in response to a letter to Chris Huhne MEP. Contrary to party policy, this letter takes an extreme pro-swpat line. Rebuttals and comments are in italics.
Letter
...It is not quite clear why there is such a belief that it is intended to lead Europe down the US path because that battle, (such as it was) was won a long time ago and the legislation both makes it clear that Europe does not intend to do the same as the US and provides a defence against further attempts to force it that way.
How can that be if current EPO practice, which is already close to the US practice, must be the baseline for a directive? So she's contradicting herself, by saying we're not going down the US path but then later that we have no choice.
In the short term it makes no difference to the software industry as in the absence of the directive the status quo will be maintained and they are essentially the same.
In the absence of the directive, most countries will not enforce EPO-granted software patents, as she herself admits later on.
(My view as a European Patent Attorney is that the directive is probably slightly more restricting than the status quo). The risk then comes if Europe gets forced in the future via the WTO to accept something that is more like the US practice
Why should the WTO force US practice on the EU? WTO != US, indeed the US seems to be showing contempt for the WTO lately. The fact that the US is using bilateral free trade agreements to impose strong-arm intellectual property laws (including software patents) on nations suggests that it does't think it can impose these laws through the WTO. The EU is probably the strongest bulwark against US economic imperialism. Why should the EU capitulate to pressure that we don't even know can be brought to bear?
Background. Patents are granted for inventions that are new, not obvious (have inventive merit) and are capable of industrial application. These may be products or processes (methods). Historically legislation has specifically excluded from patentability some fields such as mathematical methods, business methods and computer programs which at the time of their exclusion were considered entirely 'mental operations' and therefore not capable of industrial application.
i.e. before "industry" was redefined to mean practically any commercial activity. Should patentability be extended to things just because the meaning of words we use changes? That is, does the fact that we now talk about "the music industry" and the "retail industry" mean that innovations in those fields now have industrial application, and therefore should become patentable?
The problem in recent years faced by Patent Offices and courts has been how to interpret 'computer programs'. This has become highly relevant, not just because of the increasing use of computer programs but rather more because of the nature of things that are done nowadays by computer programs.
In particular there are now substantial areas of technology where software has replaced conventional analog or digital circuits. Examples range from control of mechanical systems such as an internal combustion engine to control of packet switching in network routers, which underlies communications technology.
Irrelevant. An algorithm is an algorithm is an algorithm, whether it's used for calculating a financial formula or for packet switching. It's still a mental process, and therefore should still not be patentable. If an algorithm for packet switching is patentable (even if that were desirable, which I doubt), then there is no way of stopping financial algorithms or computer-implemented business methods from being patentable.
The fact that software is replacing hardware is also irrelevant: software doesn't need heavy expensive industrial processes to manufacture. Therefore, the fact that software can now be used in place of hardware simply means that patents are no longer needed. This may be inconvenient for the manufacturing companies which are used to playing games with patents, but that is a problem for them to deal with.
Further, novel mechanical systems are, and always will be, patentable regardless of whether software is involved in them. These are completely outside the scope of this directive.
The European Patent Office adopted the practice that a 'computer-implemented invention' was more than just a 'computer program' and was patentable when there was also a technical effect because this fulfilled the fundamental requirement that a patent be 'capable of industrial application'.
I thought the "technical" and "industrial" criteria were separate. As explained below, given the way the EPO uses "technical" and "industrial", this in practice makes any software process patentable, and that is the practice that the directive in its current form imposes.
...
More recently, some European countries expressed the opinion that their courts might take a different view on 'computer implemented inventions' and deem patents that were to a method implemented by a program (as were being granted by the European Patent Office) invalid. At this stage clarification became desirable for a variety of reasons. One of these was that if the practice of the European Patent Office were not upheld in all courts, then the same logic that caused the EPO to regard these computer-implemented inventions as patentable (by virtue of being methods capable of industrial application and belonging to a field of technology) could also be used to find against the EU in respect of its TRIPS obligations (Trade Related Intellectual Property Agreements which are part of the WTO agreements) under which patent protection for all areas of technology must be provided.
In other words, we must consider the fact that not all national courts will uphold patents granted by the EPO to be a problem, and a directive is needed to ensure that they do. So the patents on this page need to be made enforceable.
Does selling something over the internet belong to a field of technology? According to the above, it does, TRIPS says that it does, and we should capitulate to this (highly questionable) interpretation of the treaty. However, the question is not whether we should patent things in all areas of technology, the question is whether anything and everything should or can be subsumed into some "field of technology" simply by virtue of the language used to describe it.
- This was only a theoretical breach, in
that it had not actually happened,
Not true, some national courts have already struck down software patents on the basis of non-patentable subject matter.
- but coupled with the
uncertainty that the pronouncements had made the Commission considered it would be best to confirm a uniform interpretation.
The Proposals. The objective is to define the practice that the EPO has established as applicable in all countries.
In other words, to make selling things over the web a patentable "invention", as the EPO has decided it is.
This would mean no change to the practice that the software industry has experienced in recent years in Europe and no following of wider US practice.
So what are the webshop patents then, if they are not similar to the US practice?
...
Amendments passed by the Parliament, some of which were proposed by the software industry, unintentionally made matters worse. For example the amendments included several that attempted to redefine 'industrial application' and 'technical effect' which are definitions that have been in use for decades with a substantial body of law behind them on a worldwide basis.
Not true, they are not defined anywhere. In particular, the TRIPS treaty which supposedly mandates the EU needs to move toward US patenting practice while saying that it isn't doing so, does not define these terms. There is, in other words, no mandate for everything to be technical. Just for things defined as "technical" to be patentable.
The new definitions may have solved a problem (that was not necessarily there)
So the webshop patents don't exist then? They're just a figment of our imagination? Or is it that actually because it makes use of some mysterious "technical" effect, selling things over the web is actually a brilliant invention?
The terms "industrial" and "technical" need to be clearly defined in order to stop what has been happening at the EPO lately, where practically anything can be made patentable simply by couching it in technical-sounding language and saying any commercial activity is an "industry" --- a usage which is current, BTW, only in the English-speaking world. The European Parliament defined 'technical" in terms of applied natural science, and "industrial" in the sense of manufacturing. These certainly exclude what needs to be excluded (i.e. software and business methods), and it's better than leaving the terms completely undefined, as proposed by the Council and Commission.
for the software industry but they also had an impact on other areas of electronics, potentially excluding from patentability a range of electronic inventions, such as those implemented through processing of signal values, for which patentability had previously never been in doubt. This would have been a disaster for the European electronics and communications industries, including many SMEs,
It is not easy to see why it should be a "disaster" for the European electronics and communications industries in particular. Most EPO patents in these fields are held by US and Japanese companies. If exclusion of these patents is a disaster, it is much more of a disaster for the US and Japanese electronics and comms than for Europeans.
Two points about patents should be made clear. First, they are not granted on the basis of nationality of applicant, or on the basis of where research leading to them was done. Second, they affect the market within the jurisdiction in which they are granted, and are completely worthless outside that jurisdiction. That is to say, an American company can apply for a patent at the EPO, and it is can be enforced only on things sold in Europe. Alternatively, a European or Asian company can apply for a patent with the USPTO, and it can be enforced only on products sold within the US. Hence, if EPO patents owned by European telcos etc become invalid, so do those owned by US and Japanese ones, while there is nothing to stop Nokia, Ericsson etc playing the patent game at the USPTO. Indeed they currently do just that, and EU patent policy won't and can't make any difference to their ability to do so.
Thus, when Microsoft threatened, as it did a few months ago, to move its R&D out of Europe if software patents do not become available within the EU, it was either bluffing, or using its European operations as a political pawn for an unrelated issue (in which case MS's European investments and jobs are unsafe anyway).
European companies can enforce US patents in the US, and have done so. For example, the UK-based patent troll Allvoice has control over the whole text-to-speech software market since it successfully enforced a US patent on the very principle of using computer algorithms to turn text into speech. Another UK-based rent-seeking outfit, the "British Technology Group", is currently suing large software and e-commerce companies, in the US, on several broad and trivial patents relating to software updates and web user identification from server logs. British Telecom's failed patent claim on hyperlinks was also prosecuted in the US. Is this the sort of communications patent Sharon Bowles wants to see valid in Europe?
Generally, patents in the electronics and communications sectors are not held to protect any R&D, but instead for strategic purposes, such as to exclude future competitors and as bargaining chips in business deals. But that is not the original purpose of the patent system, and in an open market it is not the state's job to protect business models or present incumbents.
And sure, some European software SMEs have patents. There may be various reasons, like for self-defence (so would be unnecessary if no-one has patents), or to impress investors (but investors are nowhere near as impressed by software patents as they used to be, and indeed are now beginning to be put off investing in some hi-tech fields because of the threat of patent trolls), or because their management thinks they might win from playing the patent game. So they want to make big machine guns legally available just so they can wield their peashooter?
and would have left the EU in an actual rather than just a theoretical breach of its TRIPS obligations.
It remains a theoretical breach as it has never been tested.
Now there is a 'Common Position' from the Council, which although it has not included all the wording of the amendments,
More precisely, it included none of the substantive amendments.
- has included enough to make it absolutely clear
that for patentability a computer implemented invention must have a technical effect (i.e. beyond the ordinary interactions in a computer program) and the inventive merit must relate to that technical effect.
Which means nothing if "technical effect" is not defined.
Detail from the Common Position. The Council Common Position specifically includes the amendment passed by the Parliament and supported by the Liberal Democrats that "In order to be patentable, a computer-implemented invention must be susceptible of industrial application and new and involve an inventive step. In order to involve an inventive step, a computer-implemented invention must make a technical contribution".
Not one of the better amendments, since the "inventive step" and "technical contribution" are actually two separate things.
- - A computer program as such cannot constitute a patentable
invention
Meaningless if any program which does something useful isn't a progam "as such".
- - Inventions involving computer programs, whether
expressed as source code, as object code or in any other form, which implement business, mathematical or other methods and do not produce any technical effects beyond the normal physical interactions between a program and the computer, network or other programmable apparatus in which it is run shall not be patentable.
"Normal physical interactions..." doesn't mean anything. And in any case that is not how the technical effect is claimed --- see below.
- - A claim to a computer program, either on its own or on a
carrier, shall not be allowed unless that program would, when loaded and executed in a computer, programmed computer network or other programmable apparatus, put into force a product or process claimed in the same patent application.
/The "unless" condition here is always true. All programs, except perhaps Perl poetry, put into effect some sort of product or process, and according to the directive, any such product or process is patentable (provided it has a "technical contribution" of course, but that's easy to find). So the above paragraph is deceptive. It purports to protect programmers from being sued over programs they've written which infringe someone else's software patent, but in fact it does the exact opposite. Contract programmers, in particular, should beware if a directive containing this gets passed. (Programmers who are wage-slaves are in a safer position as the employer will carry the can --- but of course they might lose their jobs as their employers decide the software business is too risky.)/
And in the accompanying explanation:
-The mere implementation of an otherwise unpatentable method on an apparatus such as a computer is not in itself sufficient to warrant a finding that a technical contribution is present.
Indeed, the "technical contribution" is something like requiring fewer mouse clicks, efficient use of screen space, or less computing time taken. For practically any computer-implemented data processing or business method, some such "technical contribution" can be found, making, in practice, any software method patentable. This is exactly the sort of nonsense the European Parliament aimed to put a stop to, and which the directive in its current form allows.
- Accordingly, a computer-implemented business method, data processing method or other method in which the only contribution to the state of the art is non-technical cannot constitute a patentable invention.
But given the broad definition of "technical" used by the EPO, that excludes nothing.
- In addition to what the directive actually covers, many have mixed in comments about particular patents (often US ones rather than European ones).
Most US software patents have also been granted by the EPO.
- US and Europe must not be mixed up, we have different laws. Unfortunately even in Europe mistakes are sometimes made about granting patents for something that turns out to be trivial. There are remedies for this and it should not be used to confuse other issues
Intellectual property lawyers tend to live in a cloudcuckooland in which laws only have the effects they are intended to have, i.e., there are no unintended consequences. Or, even if there are unintended consequences, these are problems for someone else to deal with. That is, real-world concerns have no business interfering in their cosy legal framework. The above comment is a prime example of this mentality. But when the mere threat of legal action is often enough to force people to abandon a project, when people threatened by a patent lawsuit choose to settle even when they are very likely to win, this "pass the buck" attitude is not acceptable.
The problem of trivial patents is very much a problem for patent law, which the law must deal with. If the legalists don't like that, then tough, laws need to be framed to consider their empirical effects, and patent law is no different.
Sentimental left-wingers used to say that Communism might have worked but Stalin messed it up. A counter-argument is that Communism was inevitably going to be hijacked by a Stalin, and we have an analogous situation here. Trivial patents and abuse of process are inherent problems in a software patent regime, because of the nature of software development and the cost of patent litigation. In other words, the system is unworkable. Not that I'm comparing patents with Communism, although the legalistic approach to policy taken by many intellectual property advocates is akin to that of political ideologues. And it sits rather uneasily in a party (the Liberal Democrats) which cultivates a vague suspicion of ideology. I could also add that a software patent regime would lead to a kind of corporate state, involving the government picking winners and protecting incumbents by monopoly grants.
- In general the European Patent Office aims to keep the level of inventive merit high.
Even if it does aim to do so it is not succeeding. Probably with software patents it is not possible.
