Software patents: Brussels threat to our digital freedom
|
"He who receives an idea from me, receives instruction himself without lessening mine; as he who lights his taper at mine, receives light without darkening me."
- Thomas Jefferson To the lay person, patents typically conjure up an image of an eccentric scientist toiling in his garage to 'build a better mouse trap'. It is a stereotype that advocates of strengthened patents love to promote because it keeps the illusion going that their intention is to protect the small inventor. But that couldn't be further from the truth. Most software patents are held by large corporations, predominantly from the US and Japan, according to research from the Foundation for a Free Information Infrastructure (FFII). These corporations, together with supportive patent offices are pushing forward new legislation in the European Union that copies the US model of allowing patents on 'computer implemented inventions', (i.e. computer software). Surprisingly, in spite of this support from government and business institutions, there isn't a single study that concludes that software patents are economically beneficial. Indeed, a report from the US Federal Trade Commission expresses concerns over the damaging effects of software patents in the US. The support from patent offices becomes a little easier to understand when you take into account the fact that most patent offices, including the UK's (pdf), are funded purely from patent processing fees. So a growth in the number of patents directly benefits them. Described as a minefield for the software developer by Free Software guru Richard Stallman, patents differ from copyright in that they can be accidentally infringed. Even if a developer comes up with a patented idea independently, he or she could still be sued for damages by the patent owner. The only way for an independent developer to avoid the patents 'minefield' would be to meticulously search patent files. But thats completely impractical. Every working day, programmers work through scores of ideas - and a search of any one of these could take up the whole day. It's clearly an impossible task for that solitary boffin pottering about in his garage. Large corporations, however, can avoid this burden by 'cross licensing' their patent portfolio with that of other companies: that is to say, by striking a deal allowing each to use the other's patents without worry of litigation. The irony is that patents were originally intended to promote innovation. But imagine if patents had been applied to the field of music. If the twelve bar blues had been patented back in the 1900s in the US, would we have witnessed the same melting pot of musical styles over the twentieth century that led to rock 'n' roll? It seems unlikely. And computer software evolves in the same way as music, with current ideas being combined and re-combined in innovative ways. Implications for poorer nations What makes the issue of software patenting more vital still is the powerfully inhibiting effect it can have on the developing world. The new information and communication technologies (ICTs) offer important new opportunities for economic development in poorer nations. It is vital for these nations to access the benefits of ICTs improved efficiency, processes and communication now boosting developed nations. They also present a new export opportunity, as companies in richer nations increasingly outsource software services to countries like India where skilled labour is available at lower cost. And the continuing success of Free/Libre Open Source Software (FLOSS), like the Linux operating system, is seen by many as a practical opportunity for developing countries to get a technological 'leg up', as these free systems can provide a basis for infrastructure around which services like consultancy, training, and customisation can be built. But these promising opportunities are threatened by the expansion of EU legislation on software patents. Its effects would not be immediate or direct, but in the longer term acceptance of software patents in Europe would be a key step towards a harmonised worldwide patents regime, enforced through the World Trade Organization (WTO). In this future world (which is only round the corner), small software developers, in the global south as well as the north, will find themselves faced with negotiating that patents minefield, unable to compete with the vast patent portfolios of the global software giants. The irony is that, during the 19th century, the US based much of its economic development on copying technologies from Europe - deliberately flouting the patents regime in place at the time. Call to action If we are to avoid this future scenario, now is the time to take action! The FFII is issuing an Urgent Call to European national governments.
Related links:
David Heath works for OneWorld.net as a software developer, where he builds the software which runs the OneWorld.net website. David is interested in the technology and philosophy of free/libre open source software and its role in international development. --- OneWorld Guest Editorials represent the viewpoint of the authors and not necessarily that of the OneWorld Network. |
User comments
"The Council draft is the one which will give money to lawyers"Time: 30.09.2004 15:25
Comment: Well in a sense I am FFII (I'm on its board) so I suppose I must be telling myself all this. :)
Er, "cause-effect relations" means exactly what it says --- something is put into action (the cause), producing a predictable result (the effect). "Controllable forces of nature" is also straightforward. "Controllable" in the sense that they can be explained and manipulated using the laws of science. All the things you mention (wind, tide, gravity) fall into that category. They differ from logic (which is not a force of nature) and from, for example, sociological factors (which are not controllable as we have free will --- this unambiguously prevents pure business-method patents as they rely on prescribed human behaviour to work as described). Neither of those things need many lawyers to interpret. The (real) "current position" is that there are about 30,000 EPO-granted software patents which may or may not be valid depending on jurisdiction. Certainly I agree with you on one thing: no directive at all may be preferable --- preferable, that is, to the Council draft. The Council may say they don't want to change anything, but in that case what's the point in having a directive at all? Seems a pointless waste, to spend many man-years on a law which simply says "We're not changing the law". That alone is a reason to suspect a hidden agenda. Not only that, but the fact that patent lawyers seem to be rallying to support the Council draft suggests that it is that, and not the EP draft, that would benefit lawyers more. Turkeys don't vote for Christmas, and nor do they lobby for it. FFII say they don't want to change anything, in the sense that we want to return to the original spirit of the EPC which has been corrupted by the EPO. The Council say they don't want to change anything, in the sense that they want what is patentable at the EPO to continue to be patentable (but also (here's the sting) to make sure that those patents have legal force throughout the EU). So the question is, which status quo? I suppose it would be nice if the EP were in FFII's pocket seeing as the Council and the Commission seem to be in the pockets of groups like the BSA, FICPI, EICTA etc. It would make a refreshing balance. But in fact it's not true. The EP had to be persuaded of our arguments and without such an effort, legislators are likely to default to an IPR-inflationist position. FFII also does not assume that the 2nd Reading vote is ours --- again consider the EP gene patents votes. As a contractor, you should be very concerned about the possibility of unheard-of patent owners coming out from apparently nowhere claiming exclusive ownership of some apparently straightforward programming method that you and your colleagues had been using for years. According to the Council draft, the patent holder would own your program (which you had written), and you could be sued for distributing it. Nice work for the lawyers. But you won't be able to earn your crust if it strikes you. |
"Money for lawyers"Author:
Clive Robertson
Time: 24.09.2004 11:58
Comment: I can see you really believe what FFII are telling you, but I still say that the Parliament's version is a licence to lawyers to print money. I mean, take the following test which the Parliament is proposing:
"In determining whether a given computer-implemented invention makes a technical contribution, the following test shall be used: whether it constitutes a new teaching on cause-effect relations in the use of controllable forces of natures and has an industrial application in the strict sense of the expression, in terms of both method and result." How many lawyers will it take to work out what this means? "cause-effect relations" ? As for controllable forces of nature, I thought the whole concept of "force of nature" (eg. wind, tide, gravity etc) is that they are uncontrollable. I'd be much happier sticking with the current position. The contracts keep coming in, there's work out there. If the Parliament changes all this by introducing what FFII thinks the current law should mean, I could be stuffed. At least the council is saying that it doesn't want to change anything - the Parliament seems to be in FFII's pocket and is hell-bent on changing the law to suit FFII. That won't be good for industry, and it won't be good for lots of people like me who make a living in industry. It's OK for you to say "rest assured - it won't happen", but I'd rather not have a directive at all. Then I don't have to worry about what all these strange words mean, and I can continue earning a crust. Clive Robertson |
"Software Patents"Author:
Alex Macfie
Time: 23.09.2004 10:26
Comment: The European Parliament's amendmens will not make any industrial processes or products unpatentable, regardless of whether they involve software. They simply restore the original intent of the European Patent Convention, that the inclusion of software should neither add to nor negate patentability of a process.
For example. A new type of antenna for a mobile phone ---for example, one which was extra-sensitive --- could well be patentable according to the no-swpat rules. Making antennae more sensitive or durable requires consideration of physical effects, such as the effect of signals on a receiver. This would be true even if part of the antenna included some embedded software, or it was driven by a computer program inside the mobile phone. But a 'convenience' feature on a mobile phone --- for example, making keys 'lockable' --- should not be patentable. This is done purely in software, no physical effects come into play, and it is also extremely trivial to implement. It also happens to have been done before, in ordinary computer keyboards. But locking keys on mobile phones is patented at the European Patent Office. I choose this example because one of the European Parliament's proponents of swpat, Conservative MEP Malcolm Harbour, used this as an example, appearing to believe that locking keys on a phone *should* be patentable, and accusing the anti-swpat people of opposing patents on 'new technologies' such as mobile phones. The above is an example of how proponents of software patents deliberately conflate things that should be patentable with things that definitely should not. Rest assured that a company which aims to sell products or services, as opposed to one seeking to engage in patent warfare, has nothing to fear from the European Parliament's amendments. If you think it's a "bit tough" to expect patent offices to consider all prior art when making decisions, then that is perhaps a good reason why software patents are a bad idea --- there is so much more prior art to consider than there is in most scientific fields, because writing software is a cottage industry, that the system would simply be unworkable. The fact that patents are routinely granted for programming techniques which the IT community has known about for years shows this. The outcome of the Bromcom case does at least mean that schools will not face litigation for setting up their own wireless networks with off-the-shelf equipment to transmit pupil management data, something which Bromcom (or rather, Frontline Technology) was threatening. This is common sense: it can be done with any other sort of data, so why not attendance data? The problem is not so much how the EPO will interpret the directive (it's already deliberately misinterpreting current patent law as proven by the pure software patents already granted), it's how courts will interpret it. FFII's concerns are indirectly confirmed by the statements of opponents of the European Parliament's amendments, who complain that the directive will invalidate patents already granted by the EPO, and who support the Council draft because it codifies existing EPO practice. Existing EPO practice --- patents which would be invalidated by the European Parliament --- includes patents such as those found at http://webshop.ffii.org/. I have already given an example of how "the plain meaning of the text" is not as plain as it claims to be. You are reading what you think it's supposed to mean, not what it actually means, or how lawyers could interpret it to mean. Blind faith in the judicial system to overturn bad practice is mistaken: the US got where it is thru a series of court decisions starting with Diamond v Dehr in the early 1980s and ending with the State Street decision of 1998 which opened the door to business method patents. The US legislature did not substantially change patent law during this time. The ECJ would probably strike down current EPO-granted software patents under current patent law, which is based on the EPC, which clearly states that computer programs are unpatentable. This is not certain, but it does seem that the main reason why Europe is relatively (as compared with the US) free from frivolous software patent litigaion is that patent-holders don't believe they'd win. But the judicial system can only go by existing law. The directive changes the law. The Council's definition of "computer-implemented invention": "... any invention the performance of which involves the use of a computer, computer network or other programmable apparatus and having one or more prima facie novel features which are realised wholly or partly by means of a computer program or computer programs;" encompasses any new and useful programming technique. Combined with the lack of a definition of "technical", this means that there would be no law under which software patents can clearly be struck down --- you have to find prior art, and most people faced with patent infringement suits can't afford to go there. [The Bromcom case was unusual --- an organization (the DfES) having both the resources *and* the incentive to challenge a bad patent. Most companies choose to settle because it's so expensive even to get over the first hurdle.] Ministers tend to listen to civil servants, and in patent matters the civil service tends to be dominated by people who directly benefit from more patents being granted --- patent officials and patent lawyers. Parliamentarians are less prone to this influence. Whether MEPs are "second-rate politicians" rally depends on the country --- in some the MEP role is taken much more seriously than in the UK. The problem generally may be that MEPs have so little power. |
"Software patents"Time: 20.09.2004 12:01
Comment: I'm not sure whether I'd call the Council's language clever, but the Parliament's language could certainly never be described as clever. It's diabolical. It would probably put me out of work.
I write control software for the automotive industry. It's a very competitive, if not to say cut-throat, industry. The manufacturer I work with insists that it could not justify spending loads of money (eg. paying my contract) if it could not protect its new products with patents. Apparently when they started up (before my time), they didn't bother patenting lots of things, and then they were nearly put out of business by a competitor who pinched one of their employees (so the story goes) and a couple of inventions along with him. I have been following the Bromcom case too - I have written some software for schools in the past (eg. timetabling and reporting). It looks as though the judge only invalidated two of the claims, and even then, only on the basis of a trial in a school over a decade ago. Bit tough expecting the EPO to know about that isn't it? When I described FFII's analysis as shoddy, I was referring to the way it just makes blind assertions as to how the EPO will interpret the directive. In some cases, FFII suggests that it will interpret some provisions in precisely the opposite way to the plain meaning of the text. I don't see how they could do that. The Court of Justice in Luxembourg (or wherever) would slap them down. But FFII seems to be saying it in order to scare people into lobbying their MEPs etc.. I can see how the European Parliament's amended version of the directive would be good for some progammers, but for most of us - ie. that make a living out of writing programs in industry, it could be catastrophic. That seems to be the case in the auto industry for certain, and probably in telecoms, aerospace, and transport generally. As a matter of interest, if you think Sainsbury is out of his depth on this, what makes you think the MEPs in Brussels are any different? I'll admit to being a Eurosceptic, but even those who are pro-Europe admit that MEPs are second-rate politicians. (The third-rate politicians end up in the Scottish Parliament, or the Welsh Assembly so I'm told.) |
"Software patents: Brussels threat to our digital freedom"Author:
Alex Macfie
Time: 18.09.2004 06:04
Comment: Clive
You seem willing to be deceived by the clever language of the Council. But the EU has been here before. In the late 1990s MEPs voted against gene patents in first reading, but then broadly supported the Council's "compromise" in second reading. The EU now has a directive on gene patenting which is considered by all sides to be probably the most patent-permissive in the world, and even major business players have concerns about it on that score. Look more closely at Article 4, para 2 of the Council text. It starts with "A computer-implemented invention shall not be regarded as making a technical contribution merely because it involves the use of a computer, network or other programmable apparatus." But that is not how computer-implemented "inventions" are claimed. The claimed "technical contribution" is the physical processes within the computer apparatus which allow the computer program to run, or the claimed efficiency which results from the computer being used, or any other useful result of a program, eg more efficient use of screen space, or the need for fewer key presses or mouse-clicks. The UK office doesn't seem to interpret this quite as broadly as the EPO ---- but it still allows many dumb patents, such as the Bromcom/Frontline pupil management by wireless networking in schools patent which the Dept of Education challenged in court. Luckily a school had previously implemented a pupil management system in, and when that was submitted as prior art the patent was largely overturned (the use of wireless networking being considered 'obvious' in the light of that system). [But what if prior art hadn't been found? Taxpayers money would now have to be wasted on "licencing" a stupid patent.] I know what you're going to say. Programs which "do not produce any technical effects beyond the normal physical interactions between a program and the computer ... shall not be patentable." So that's OK then. But nowhere is the phrase " normal physical interactions between a program and the computer" defined, and it has never been previously defined in caselaw. So patent offices could interpret it however they wish. This is indeed a good exmple of "Brussels fog". But we can't afford to shrug it off like you do. A law on patentability needs to be *crystal-clear* to excluse exactly what neds to be excluded from patenting. The "fog" isn't accidental --- it's there deliberately to allow patent ofices leeway to make clever interpretations allowing patents on pure software. For example, since "technical" is nowhere defined, what's to stop the sort of interpretations I mentioned above? "Industrial" is not defined either. ["One-click shopping" is patentable because it produces the "technical" effect of reduced mouse-clicks, and is applicable in the retaiul "industry".] Some parts of the Council document are just plain deceptive. Look at Art 5 para 2: "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 in accordance with paragraph 1." Looks like it disallows claims on computer programs. But look at the "unless..." condition. ALL programs of any utility put into force some sort of "product or process". The only programs that wouldn't is Perl poetry. So the "unless" condition is ALWAYS TRUE. So the above paragraph really means: "A claim to a computer program, either on its own or on a carrier, shall be allowed." In some ways the Council text forces national courts to follow the EPO doctrine, by overriding national case-law which clearly exclused software from patentability. That might mean that even the UK Patent Office, which is among the more patent-permissive national offices in the EU, would be forced to become even more pro-patent than it already is. How is FFII's analysis "shoddy"? What is its "dodgy" information? It is based on a simple fact: After the European Parliamentary vote last September, the Council discarded nearly all of the EP's amendments which would have really limited patentability, and returned to the original deceptive language. Far from "shooting itself in the fool", it would be very strange to support *this* document when it is so similar to the one that FFII fought so strongly against before the European Parliament amended it. Perhaps you ought to take a look at the European Parliament's version of the directive, so you can see what a law that *really* limits patentability looks like: http://lwn.net/Articles/50722/ You're naive if you think that ministers always read *and understand* what they're signing. Britain in particular has a strontg "trust the civil service" culture, and a political establishment which tends to think that "intellectual property" is always a good thing. Sainsbury believes the assurances from the Patent Office, and doesn't want to look any further. This is quite normal behaviour on issues which are considered by the political establishment to be "uncontoversial". So I can easily believe that Sainsbury signed the letters without really understanding them. In summary: "Don't believe the Council hype" Read the Council document carefully. Compare it with the European Parliament's version. Decide which you prefer: one with fuzzy, easily-circumventible pseudo-limits on software patentability or one with crystal-clear limits which cannot easily be circumvented. |
"Software patents"Author:
Clive Robertson
Time: 17.09.2004 10:03
Comment: But the 30 or so rejections are just the tip of the iceberg. Aren't they just the ones where the applicant appealed against the Office's refusal notice?
I looked at the document (thanks for the link David - however did you find it?). It doesn't seem to be very clear in places - typical Brussels fog. But it does clearly say in article 4a paragraph 2 that merely using a computer etc shall not be regarded as making a technical contribution. Alex says that the EPO considers even the fact that a computer is used is enough to constitute a technical contribution. Seems to me they wouldn't be able to continue granting patents for stuff like this if the directive is passed. Is is possible that the EU is trying to rein in the EPO, and stop them from making the same mistake as the Americans? The government responses may be ghost-written by the Patent Office, but all the ones I've had were signed by the minister. Not pp'ed - he actually signed it himself. Can't believe he didn't read it first. FFII's analysis looks rather shoddy to me. Sorry to be critical - I know they're trying their best. But they seem to have based their entire campaign on dodgy information. Having declared all-out war on the establishment in general, the leaders would look stupid if they backed down now and said "Ooops -we got it wrong". But having read the proposed legislation, it looks to me as though it is intended to restrain the Patent Offices from drifting any further towards the American shambles. Surely the FFII is shooting itself in the foot by wrecking this directive. Clive Robertson (Independent programmer) |
"Software Patents"Author:
Alex Macfie
Time: 16.09.2004 14:23
Comment: I wouldn't say a list of 30 patent rejections is "loads" given that patent offices review thousands of applications every year and grant thousands of patents.
Saying there isn't a problem of patent inflation because some applications are rejected is like saying grade inflation isn't a problem because some students are still failed. If you believe that student grades are being inflated you will undoubtedly accept that some students are still being failed for the sake of appearances to show that some procedure is still being followed, and much the same situation applies here. The point is that lowering the standards, or broadening the definition of patentability, means that many patents are being allowed which should not be. I quickly read a few of the ones linked to. Not all relate to software, but those that do tend to mention this magic formula "technical contribution", which is the thin thread keeping us from American practice. The "technical contribution" can be practically any useful effect caused by software, eg more efficient use of memory or screen space. This allows most software patents, but the ones rejected didn't even have that. Some of these patents would probably stil be accepted by the EPO, which considers even the fact that computer apparatus is used to constitute a "technical contribution". As for the gvt's claim that its proposal would not introduce software patents, the British gvt respeonses are ghost-written by the Patent office. 'nuff said. |
"Examples of worrying patents"Author:
David Heath
Time: 15.09.2004 17:43
Comment: Hi Clive,
thanks for your comments. Whilst I am sure that some software patents may be rejected (and practice probably differs between the UK patent office and the European Patent Office), there are still many worrying software patents granted. For some examples, I would recommend taking a look at the following links: Patented European webshop European Software Patent Horror Gallery Regarding the actual text of the proposed legislation, it is available here (pdf format). You might also like to refer to FFII's very detailed analysis of the document. David Heath |
"Software patents"Author:
Clive Robertson
Time: 15.09.2004 09:05
Comment: If the UK Patent Office is so keen to follow the US Patent and Trade Mark Office and grant patents for computer software, how come they seem to be refusing so many applications in this area?
I've just been browsing the UK PO's website, and there are loads of rejected applications listed there (link). I've had a couple of replies from the "responsible minister" via my MP, and the government seems to be adamant that the proposed directive will not change the current legal position here in Europe. But FFII says it will introduce US style patents. Someone must be lying. Where can I see the text of this directive so that I can see for myself which side is telling porkies? Clive (Independent programmer) |
Printable version
top of the page