EU software patents: the readers speak

On Monday, we wrote a fairly lengthy piece on the upcoming vote in the European Parliament over whether to extend patent law to computer software. We got some interesting responses. And, considering the underlying theme of the article was that open-source advocates had damaged their own case by behaving petulantly and failing to see the way the world works, there were surprisingly few abusive ones - and no flames.

One reader has taken it upon himself to send a link to the article to all his local MEPs and ask them to vote against adopting the report. There is still a chance the arguments made against extending patent law will persuade enough MEPs to vote against the measure as the vote is now scheduled for 1 September.

We originally said it was due on Monday 30 June, and indeed it was. However, at virtually the last minute, this rushed-through vote was rejected and put back to September. So thankyou to those readers that quickly pointed this out so we could change the details straight away.

Almost everyone agreed that demonising the chairwoman of the Legal Affairs and Internal Market Committee, Arlene McCarthy, was counterproductive to the open source community's aims. However, not everyone was willing to let it go and a few emails agreed it was a bad idea and in the same breath decried her as some kind of Anti-Christ. And no, there is no known relation between her and this writer.

There is no denying the animosity built up between the anti-patent lobby and Ms McCarthy is probably too great to ever be effectively dealt with. Bernhard Kaindl, who has been at the forefront of the anti-patent crew, supplied links to transcripts and audio recordings of McCarthy speaking which demonstrate she has made up her mind about the issue.

Again though, it is the job of politicians to make up their minds. It seems people are angry she made up her mind the "wrong" way. Ms McCarthy is clearly irritated by sharp criticism aimed at her but her rejection of others' amendments is quite lucid: "I will not be accepting amendments that attempt to radically redefine patent law - and I have to say not because I don't agree with a redefinition - but the reality is that if there are people that want a radical change and rewriting, I think the burden of proof is on them to explain what will be the outcome of such a radical redefinition of patentability."

She continues: "I want to be very clear that in some amendments - which I believe are a little bit confused - people have confused software invention with overall software. We are not here patenting software per se, it is very clear from amendments I have tabled that I have excluded algorithms, I have excluded business methods…"

She goes on to say she has produced compromises where they are possible but that some suggested amendments go against her particular approach and so cannot be included. You can find all the audio files here.

What about the legal aspects? Well, as the article warned, the issue is extremely complex and in simplifying some aspects of it, a small degree of misleading inaccuracy crept in. It makes no difference to the article itself, but for the sake of completion here are two emails from experts in patent law.

The granting of new patents had been simplified to the issue of an "inventive step". Ted Chwu had this to say: "In order for any invention to be granted a patent there has to be shown that the claimed invention has novelty and inventive step over the prior art. The added hurdle for software patents is that the software, when run on a computer, has to produce a technical contribution over that merely achievable by merely running any programme on a computer.

"It's a subtle test, and one way to think of it is that you have to show that you are not merely computerising tasks, but by computerising specific tasks you are producing synergistic effects."

That's a good clarification. This one is slightly more complicated. Thanks to Gary Lea: "One or two minor details to point out:

1. The EU Directive will not amend the European Patent Convention (EPC) because the EPC is not an EU treaty and the EPO is not an EU institution (though, confusingly, when the Community Patent Regulation comes into force, it is likely to become the EU's Patent Office under a contracting-out deal being negotiated by the Commission).
2. The EPC is a pan-European Convention whose signatories include Switzerland and Turkey (def. not EU Member States) which allows applicants to file a single application; this is processed centrally by the EPO and, if approved, a grant notification is sent to each designated signatory state and a "bundle" of patents is granted. The Community Patent, when it comes on-stream, will be a single unitary patent grant taking effect across all EU Member States.
3. The draft Directive is something different again - an EU Internal Market harmonisation measure (like the Copyright in the Information Society, Database and Copyright in Computer Programs Directives) which will affect the *domestic* patent laws of all EU Members States (EUMSs - all of whom, it is true, are also EPC signatories).
4. Inside the Commission, the original proposal was touted as a tidying up/belt and braces exercise to ensure that patents are treated within EUMSs in a fashion identical to the standards developed by the EPO (these being felt to be right for the EU to adopt because they are, in a general sense, European and at least they retain some limited restrictions: switching to a straight "open door" US-style system was ruled out early on). Some EUMSs, rather stupidly, apply slightly different rules to patents that they are obliged to grant under the EPC than those that go through their own purely domestic systems.
5. In reality, the present draft Directive is still actually somewhat more restrictive than current EPO practice so we may still have some tangles over which standards the EUMSs will apply.
6. Ultimately, it may be that the EU is forced to go down the route of a full-blown US-style "open door" policy anyway: the US could always decide to launch a World Trade Organisation reference on the back of Art. 27, TRIPS (IBM has already used the argument about Art. 27 on the EPO, forcing the patents door open slightly more in 1997)."

Now I bet you never wanted to know so much about patents, did you? But on to the good stuff.

In a quick explanation as to why patents could be a good thing, we argued that drugs were a great example. Billions in R&D for one drug to pop out the end that was saleable. Without patents, this R&D would never happen and hence no new drugs. People dispute this:

Glenn Thorpe argues: "I must be completely and utterly stupid, cause I think there would still be new drugs, that there would still be research and that there would still be money made. Thanks for your time." Well, you said it Glenn.

Tom Poe goes a more John Lennon route however ("Imagine there's no patents, it's easy if you try"): "Suppose, for sake of discussion, that patents do not exist in the form they are, today. A government does the research - they can at a fraction of the cost the industry insists it does - and contracts with manufacturing firms to produce the product. Seems reasonable. In fact, however, the pharmaceutical industry (juggernaut) continues to perpetuate the myth about cost of R&D. Result? Governments pay to have research done, and company manufactures product, and realises increased profits at expense of taxpayers."

Msafrin goes along similar pie-in-the-sky lines: "The fact of the matter is, the people who actually do research are scientists, who make a salary. I don't see how awarding a patent to the company the scientists work for is any incentive for the scientists to do better work. If these scientists were working for, say, a university, making the same salary, the quality of their research would be just as high (even higher, most likely). What we need is more public funding for public research, and with that we might not need pharmaceutical patents at all either. And if we did that, we'd probably have a lot less money wasted on either frivolous drugs (at least less important than certain other drugs) and marketing."

So, with the quality of argument against the drug patent example as listed above, we'll just leave it as read that patents work in some cases, shall we? Likewise the peculiar emails that argued the points made about politicians being self-interested and arrogant weren't necessarily true. These people will either soon land back to earth with a nasty bump or become one themselves and scare us all in 10 years' time.

Niall Mac Caughey had an interesting point: "As someone who has the misfortune to deal with EU legislation on an almost daily basis in the course of my job, it strikes me that an important argument against software patenting has not been addressed, at least not in so many words. I'm referring to the buzzword(s) 'anti-competitive'.

"I work for a regulatory agency. One of the fundamental principles behind our office is that of improving the goods and services offered to the public by increasing competition in the marketplace. Of course increasing competition usually means lowering prices and increasing availability, which is not quite the same thing as improving goods & services; but that's an argument for a different day.

"One of the key elements of improving competition applied by this, and a plethora of other agencies, is reducing the barriers to entry for new competitors. I believe that a cogent argument could be made that patenting software raises those barriers. More importantly though for those who believe that software patenting is a Bad Thing, is an understanding of the emotive aspects of the phrase 'anti-competitive' in the mind of your average Eurocrat. In the course of drafting our own legislation and transposing EU directives we have learned the power of this phase and how carefully we must avoid anything that smacks of this, if we wish to make a point. Although the elected representatives may depend upon the largesse of corporations for the operation of their parties, they are loath to oppose the creed of improved competition. Significantly this creed also pervades the ranks of the EU civil servants who actually craft the legislation."

What Niall is saying, albeit in a long-winded way, is that the anti-patent lobby concentrated on making patent extension synonymous with being "anti-competitive" their case would become stronger instantly.

Maarten Bestevaar has an interesting take on things: "You paint a very dismal picture. As a small software company owner, I am worried. Software is not something that you should be able to get a patent on. Writing software is not 'invention' in any sense of the word. Writing software is stringing together computer instructions. Writing a program is not 'invention'. Granted, you could 'invent' a complex mathematical algorithm and them implement that algorithm in software. In such a case, the algorithm should be patented, not the software.

"To carry the idea of software patenting to its ridiculous conclusion: If you can patent a sequence of computer instructions, you should be able to patent a string of anything else, such as a particular grouping of plants in your garden, a floral arrangement, or a string of words. I can patent 'read my lips' or any other phrase. Maybe I can patent 'forward human motion', the act of placing one foot in front of the other and then shifting your weight to the forward foot, and then sue anyone using that method..."

He ends: "If these are American values, I don't want them. Europe should seek it's own path, and not follow the US, or give in to their pressures. We have our own values, and we do not need the US."

Selva argues that patenting needn't adversely affect the open source community at all: "I think you have based your article on faulty presumptions. You seem to automatically assume that the main reason open source has been so successful against the backdrop of an monopoly powerhouse is because of the open source nature of the development process. However, this remains to be proven. My feeling is that the reason it has been successful thus far has less to do with open source nature of the process and more to do simply with the fact that it is a pervasive phenomenon without a single focus for MS to focus their efforts on 'cutting off air supply'.

"Even if patents were enforced more strictly in software, I don't think you can automatically assume this will be bad for open source community. Rather, the effect of the new rules may change the way software companies pursue open source projects but probably will not discourage it. In fact, one possible outcome may be that companies investing in Linux-based projects may finally have the incentive to start funding their own future projects by charging higher margins from licensing fees for certain aspects of their work while contributing other components under open source license. In other words, the new software patent rules may actually change the way OSS operates for the better and result in stronger, more rapidly advancing products that can better take on monopolies such as MS."

It's an interesting point. Even it does steamroller the whole philosophy of open source.

Chris Sabat has clearly made up his mind: "Patenting software is not a good idea. It stifles innovation and such. Suppose Microsoft patented Word back in the day, we wouldn't have *any* competing word processors. Stupid example but you get my point. Algorithms, this is where research goes this is where new ideas are born for doing things with data in new and interesting ways. The applications themselves are just the packaging."

Except of course algorithms have been purposefully written out of the report. The reason why is too tedious to go into now, but let's just say those people saying this is a good idea would be arguing against it if all the issues were explained. Again, patents are not exactly black and white.

Jay Stuffunk, which we suspect isn't his real name, had this to say: "The point that needs to be considered with software patents is the length of time patents should exist. Twenty years is an eternity in computer software. If you study past trends, five to seven years for a software patent is way more than enough time for a product that can be brought to a worldwide market in minutes rather rather than months or years. Any longer than that is counter productive. If you again study past trends, old outdated software IP has been used to stiffle competition and innovation. Software are tools not written works, even if they look to be written documents."

It's a fair point. But again we have accidentally strayed into another confusing area purposefully left out of the article to avoid getting bogged down: copyright. Copyright gives the authors of software immediate legal rights over what they have created. Many people argue that copyright itself is sufficient to protect software and it's a pretty strong argument. People would not be able to copy chunks of code because they would be infringing someone else's copyright. So why do we need software patents at all? The same three arguments: innovation, investment, harmonisation of laws. Patents are stronger and more in your face than copyright, meaning that companies can scare competitors off their patch and so make more money. Companies like this.

There were lots more emails - which were all read - but frankly this is too long already so we'll end with Tim Taylor's succinct summary: "I guess the politicians have two choices. Neither are related to the question being posed. The answer is 'do nothing' or 'do what someone else has done, but slightly different'. Choice 2 has to be popular because putting 'We did nothing' on a manifesto doesn't get many votes. (although perhaps it should)." ®

Related link
Open source prepares to kiss EU patent ass goodbye