Jeden z naszych czytelników (nie przedstawił się, niestety) przetłumaczył na język angielski tekst poświęcony patentom.

Sam zastrzega, że tłumaczenie nie jest idealne – poprawki prosimy przesyłać na maila – lecz może warto podesłać ten tekst również obcojęzycznym znajomym?

Dziękujemy za dobrą robotę… i czekamy na ujawnienie sie autora tłumaczenia. (dziękujemy również osobie, która prosiła o anonimowość, a poprawiła pierwotny tekst)

On the 1st of September the European Parliament is due to discuss the possibility of introducing new kinds of patents in the European Union. If the idea gets accepted by most of the deputies, soon it will be impossible to create computer programs without prior paying to the owners of patents covering the programming solutions to be used in the program.

The introduction of possibility of patenting algorithms is a significant step backwards for all of us – it will do harm to small and medium companies, as well as end-users of software. Companies will suffer because the introduction of law similar to that in the US will force them to buy expensive licenses for– patents (those being very often absurd), while the users will be harmed mostly by the domination of monopolies on the market. Producers of cheap or free software will be denied the purpose and means for further existence.

The planned changes would violate the 1973 Munich Convention (signed by 27 countries, many of them not being members of the EU; Poland did not sign the Convention). Article 52 lists types of discoveries that are not considered to be inventions, and therefore are not subject to patenting. First on this list are scientific discoveries and theories, and mathematical methods.

It’s a logical assumption. Our civilization would not develop without discoveries, scientific theories, and mathematical methods. Imagine a situation where the the famous E=mc^2 equation would be patented. In everyday life we constantly encounter objects, devices and ideas, that couldn’t be created without science or advanced mathematics.

The case with computer science is very similar. Even the simplest program accomplishes some task – it has an algorithm according to which it works. Imagine that there is a possibility of patenting every inventive idea. Several years ago such an idea was the invention of hyper links – underlined fragments of WWW pages, that move us to an other page upon being clicked. If the author of that revolutionary idea had chosen to patent it (he was able to do so, as he comes from the US, where such solutions can be patented), Internet would look totally different. The vast majority of Web pages would be rendered useless – obviously a lot of people would not be able to afford necessary fees for the usage of hyper links. The fortunate more ones would have to share their profits coming from web pages with the patent holder.

Recently, British Telecommunications (BT) tried to prove that they hold rights to a patent like this. If they happened to get a court approval for their claims, they would be able to enforce that patent on authors of every web page placed on US servers.

Another type of danger lies in patents on things often considered “obvious”. A patent clerk is not a trained scientist, and isn’t always able to recognize every nuance of a patent proposal, which is usually formed in a way giving it as wide possibilities of usage as possible.

The patent office’s ignorance and lack of awareness of widely used technologies led to the introduction of one of the most ridiculous American patents. The popular Internet bookshop was given a patent on a system called „one-click-shopping”. It allows a customer of the online shop to enter his data once and for all to the shop’s database. Later on, the customer isn’t required to input all of that data again during his next shopping, because it is already stored. In a nutshell, a patent has been given for a solution that was widely used in many Internet shops existing at that time and implemented by hundreds of programmers around the globe.

Cases like the one described above often take place outside of the world of advanced technology: sometimes a patent is given for something known for centuries. In 2001, an Australian freelance patent lawyer known by the name of John Keogh was given the rights to a „circular transportation facilitation device”. More commonly known as the wheel.

What if the European patent office proves to be less tolerant then the Australian one?

First of all, we may encounter the necessity of exchanging the patents – if an idea or algorithm is patented in the US, then that patent will be valid in the EU as well. This puts European software vendors and producers in a very awkward position, mostly because thousands of popular algorithms have been patented in the US – from the simple ones, like „one-click-shopping”, to the more advanced ones – like MP3 coding for instance.

The possibility of patent rights execution is a serious threat. Numerous useful programs have disappeared from the market in the past, because their authors couldn’t afford to pay the necessary fees.

The IT market does not resemble the „real” world, where a product includes only a few patented technologies, paid for by the end customer.

An advanced computer program uses hundreds, sometimes even thousands of advanced procedures, each and every one of them designed and written to accomplish a different task. What makes things worse, the vast majority of those could have already been patented by some company that will demand profit (ie. 1% of overall profit) for using the patented solution. If there were only a few algorithms like that – commercial software producers would deal with that without any significant problems. However, if you multiply the costs by the level of complication of the software, you clearly see that the more complicated the program is, the larger part of its profits is consumed by licensing fees. Moreover, we might find ourselves at a point where releasing excellent software fulfilling every wish of the customer can render its producer bankrupt!

The above examples have omitted the authors of so called free software – programs that are freely and publicly available with their full source code, allowing anyone to modify them according to their needs., Those programs are protected by a license allowing their free usage and redistribution. Some well known examples of such software are: the Linux operating system, the office suite, and the Mozilla Internet browser. If the creators of any of the products mentioned above considered the possibility of violating patents, none of these projects would even pass the stage of early planning.

One more type of costs wasn’t described yet – the need to evaluate whether the code currently created code doesn’t infringe other companies’ patents. This means the need to employ qualified specialists, who will track the entire code (obviously they have to be good programmers), as well as the rapidly changing patent market. In the times of merciless competition, when anyone conducting some economic activity is catching onto every method of reducing costs of operation, those obligations will cause bankruptcy of most small software companies.

Even if we assume that one would be able to find people well aware of what’s happening in patent offices, and in the computers of employed programmers, not much will change, since patent proposals can be made in confidentiality, and remain in this state until lawful protection begins to take place. A similar case could have been observed in the ’80s in the US, the LZW compression algorithm was patented. The patent proposal was waiting for a few months in the patent office. Meanwhile several programs using the method have been created. When the patent was granted, their creators were threatened by lawsuits, and the software became illegal.

How can one deal with those threats? The creators of software have three options: omitting the problem by not using the patented techniques, purchasing a license, or submitting the case to the court.

The first method doesn’t always apply. Sometimes it would mean that the program would lose one unimportant feature. In other cases the program might work much slower, because a
faster algorithm for doing the same job is already patented. Finally – it sometimes would be necessary to cancel the creation of the program altogether. Until 1997 a situation like this occurred with the idea of encryption using public, and private key – due to the patent law a monopoly could be established in using the now-popular technique.

The second option doesn’t leave us with more possibilities. First, the owner of the patent may choose not to grant a licence. Such cases would take place if the patent holder creates software similar to ours, or dislikes our model of operation (this is the problem that is faced by the creators of free software). Even if the patent holder agrees to sell us a license, he may demand – say – 5% of our profits. It’s not a high price, but remember what has been mentioned before – our program may have to use plenty of patented algorithms.

The last solution – trying to prove that given algorithm is nothing innovative. It’s not easy in Polish reality, where even the simplest cases can take years before verdict, and our country is not the only one with similar problems.

As proved, little companies can face most dire problems in the world in which algorithms can be patented. Giants, on the other hand do not face similar problems. They have three alternatives: they buy the license for the patent, they buy the patent holder, or use the opportunity of relicensing.

Large corporations (ie. IBM) possess thousands of patents. It’s extremely easy to break one of them. Let us not forget, however, that if one company has numerous patents, so has the other.. They just agree to use each other’s solutions. Problem solved.

One might ask, why should software be treated in any other way than for example industry? Why should industrial patents be tolerated, and the ones in IT not?

The existing patent system has been developed with protection of given products in mind. A pharmaceutical company could patent aspirin, a chemical one – a new type of glue. Their products would be protected – no one would be able to produce the same thing that they do without their permission. However, another company would be able to produce another type of glue. It works in the same way, but is produced in a completely different manner, with different types of ingredients used.

Allowing to put patent rights on algorithms is to the same as allowing to give someone the rights to the glue as well as to every other thing that is glued. No one would be allowed to produce any type of glue or even a glued chest of drawers without buying a license. Later on, we might learn that someone has a patent for black paint, and someone else – for drawers… A producer making kitchen tables would probably never be able to legally introduce his product to the market.
Archiwalny news dodany przez użytkownika: honey.
Kliknij tutaj by zobaczyć archiwalne komentarze.

Oznaczone jako → 
Share →