Protests Delay European Software Patent Vote 316
vinsci writes "According to CNET News, 'The European Parliament has delayed voting on a controversial software-patents directive, following protests and criticism by computer scientists and economists.' ZDNet UK adds: 'Warnings that a controversial directive could devastate European software businesses have struck a chord with MEPs. The European Parliament has delayed voting on a controversial software-patents directive... the vote, originally planned for Monday, will now take place at a plenary session starting on 22 September.' Wired also has a story on the protests."
Still... (Score:5, Interesting)
Good thing or bad thing? (Score:5, Interesting)
Don't assume (Score:5, Interesting)
Really that bad (Score:2, Interesting)
1-In order to be patentable, inventions in general and computer-implemented inventions in particular must be susceptible of industrial application, new and involve an inventive step. In order to involve an inventive step, computer-implemented inventions should make a technical contribution to the state of the art.
2-Accordingly, even though a computer-implemented invention belongs by virtue of its very nature to a field of technology, it is important to make it clear that where an invention does not make a technical contribution to the state of the art, as would be the case, for example, where its specific contribution lacks a technical character, the invention will lack an inventive step and thus will not be patentable.
3-Furthermore, an algorithm is inherently non-technical and therefore cannot constitute a technical invention. Nonetheless, a method involving the use of an algorithm might be patentable provided that the method is used to solve a technical problem. However, any patent granted for such a method would not monopolise the algorithm itself or its use in contexts not foreseen in the patent.
4-This recital makes it clear that it is not enough to specify the use of a computer (i.e. of technical means) to make a computer-implemented invention patentable. The invention as a whole must make a technical contribution. Ordinary data processing is not enough.
5-It is important to clarify that not all computer-implemented inventions are necessarily patentable. However, computer-implemented inventions should not be excluded from patentability on the sole ground that they specify the use of a computer program. By stressing the fact that a patentable computer-implemented invention, albeit belonging to a field of technology, must make a technical contribution to the state of the art and by drawing attention to the problem and solution approach used by the patent examiners at the European Patent Office in assessing inventive step, it is intended to avoid allowing inventive but non-technical methods (including business methods) to be regarded as making a technical contribution and hence as patentable merely because they are implemented on a computer.
Well it doesn't seem so dangerous
This bill needs to fail (Score:4, Interesting)
What it actually means to them is that the Yanks will bully them into passing laws similar to the DMCA and their obscene IP protection laws.
Since NZ (where I live) is discussing trying to get an FTA with the US too, I hope the Aussies tell the US where to stick their restrictive and absurd IP laws.
If this bill fails, it's easier for other countries to tell the US that their laws are so stupid that the only people using them are themselves. If the EU folds, then the result of the world just becomes a row for xxAA to bulldoze with the support of the US government.
Re:The US will put massive pressure on MEPs (Score:3, Interesting)
However, I'm not aware of ANY lawsuits concerning any of those software patents. IMHO nobody is crazy enough to sue someone else, because chances are the courts would just invalidate the patent (because it contains just non-patentable stuff according to the law). Remember, the patent office GRANTS the patent, but the COURT decides if it is VALID.
Let EU be beginning (Score:5, Interesting)
Re:Yay for Europe! (Score:3, Interesting)
I wonder if this same broad type of protest would have some effect on US congressman and the like, if it was coupled with a massive email/mail movement informing congressmen of the reasons behind the protest and the problems with current patent law. With sites that have a very wide range of users they effect such as Apache participating, I can't help but think that it would make some people take notice and listen.
Of course, all this may have been tried before, and failed before. I just can't remember ever seeing so many open source web sites shut down in protest like they have been over this issue.
Re:Actually, here's how it is: (Score:5, Interesting)
Thanks for these uplifting news.
It's much more complex (Score:5, Interesting)
Patentability of hardware is a well accepted principal. But what if the hardware contains software? For instance when it's a cellphone. Should the hardware still be patentable? What if the phone is only special because of functionality implemented in the software portion? For instance if it can talk to a Jabber server. Should it then be patentable because of the special features?
Now what if it's not Jabber, but some other IM server and the intelligence is in the server. Should the system of phone plus IM server be patentable? If not why not? If yes, then aren't you allowing patenting of software on general purpose hardware?
So that's the background.
Having organised a couple of meetings on the issue in Cambridge, I'm of the opinion that the case that the directive is damaging is overstated. The author of the language of the amendments introduced in JURI says that by the nature of the EU system the language can't be very tight, but that a key feature is that patentability can now be reviewed by the courts, and that JURI has made its wishes clear in the Recital, which courts use as a guide to the intention.
There is still a crucial issue of how Free and Open Source software authors are protected. The directive is inadequate in this regard, but then the situation as it stands is inadequate. We need to take a social decision that protects authors of Free and Open software because they make the efforts of their labours available without charge and that's to the benefit of society.
An interesting side effect of such a settlement would most likely be a decrease in software patenting in favour of the use of trade secret. This isn't necessarily a god thing; patents were invented to make it possible to expose trade secrets in return for a limited monopoly on their use. An example - if you invent the ultimate search algorithm but kept as a trade secret it might never enter the public domain.
A second serious problem is the length of a patent - around 20 years. For software, which typically has a life span of 5 to 8 years, this is ludicrous. On the other hand it typically takes around 2 to 4 years for a patent to be granted. But software intro cycles are around 12 to 18 months. So unless you have a spectacularly good invention, or some indirect need, it may very well not be worthwhile patenting. (The number of patents involving software suggest that this isn't generally true - numbers of 15,000 to 30,000 in Europe are commonly reported.)
The way to solve the problem of software falling under the same banner as hardware is to alter the European Patent Convention to vary the rules for software. This won't be easy, but it's probably possible.
Jeff Veit
Re:Let EU be beginning (Score:4, Interesting)
Software cannot be patented in Canada already (Score:5, Interesting)
"The difficulty with software is that programs generally centre around the use of mathematics and algorithms. It is clear that a software-related invention must do more than merely perform a calculation and must be more than an algorithm embodied in software. Otherwise, it will fall into the category of a "mere scientific principle or abstract theorem" and therefore be unpatentable.
The test for determining whether a software related invention is patentable was set out in Schlumberger Canada Ltd. v. Commissioner of Patents (1981), 56 C.P.R.(2d) 204 (F.C.A.). The test considers, what (if anything), according to the application, has been discovered. The court held that if the invention was merely the discovery that by making certain calculations according to certain formulae, useful information could be extracted from certain measurements, then the application should be refused on the basis that it lacked patentable subject matter under section 2 of the Patent Act. Essentially, the court suggested that if the program interacted with physical objects other than the computer in a novel, useful and inventive way, the invention was patentable. In response to Schlumberger and related decisions, the Patent Office developed new guidelines, used by the Patent Office Examiners to assess the patentability of computer related subject matter. The guidelines are as follows:
1.Computer programs per se are not patentable;
2.Processes which are unapplied mathematical calculations, even if expressed in words rather than in mathematical symbols, are not patentable;
3.A process and/or computer program which merely produces information for mental interpretation by a human being is not patentable, nor does the process or program confer novelty upon the apparatus which uses it;
4.Claims drawn up in terms of means plus function which merely produce intellectual data are not patentable;
5.New and useful processes incorporating a computer program, and apparatus incorporating a programmed computer, are directed to patentable subject matter if the computer related matter has been integrated with another practical system that falls within an area which is traditionally patentable; and
6.The presence of a programmed general purpose computer or a program for such a computer does not lend patentability to, nor subtract patentability from, an apparatus or process."
Thus, if a computer program is used to interact with a system or as part of a process, ie. electronically controlled sawmill machinery, then the software could be patented. Which raises the question on how the LZW algorithm was granted a patent in Canada (patent 1223965) as it clearly is nothing more than an algorithm.
Re:Only 400 Protestors? (Score:5, Interesting)
It's especially telling that the pro-swpat lobby did very little effort to refute our arguments or even to simply discredit us (e.g. calling us zealots that only want all software to be free/gratis). There were some weak attempts, but very few.
Re:Non-exclusive software patents? (Score:3, Interesting)
Not having software patents in Europe, but having them in the US is a strategic advantage for European companies. They can then even take on IBM, just like Eolas took on Microsoft (Eolas doesn't develop anything itself, so Microsoft couldn't find any of its own patents on which Eolas could infringe). I'm not advocating this kind of business strategy (in fact, I am strongly against this kind of leeching), but it may help in getting rid of swpat in the US as well.
Keep in mind it's mainly the BSA that's pushing for software patents in Europe (a leaked word document with one of the first drafts of the directive, had the name of the European head of the BSA in the author field).
Re:It's much more complex (Score:4, Interesting)
The European Patent convention explicitly states that computer programs as such are not patentable. This means that adding software to something does not make something suddenly unpatentable. The opponents of software patents are not asking for this to be changed.
No, because they're entirely realised in software. Otherwise, since the phone has a built-in processor that's simply executing some code which allows it to talk to a Jabber service, you would just have patented the ability of talking to Jabber services on any kind of computer.
There is no inherent difference between programming a cpu that's inside a cell phone and one that's inside a desktop PC. Yes, you have less memory and speed available, but if that's a criterium, then any kind of software improvement that causes less memory to be used or more computing efficiency would have to be patentable.
I don't think so. Have you looked at how http://swpat.ffii.org/players/ibm/#gajn [ffii.org] in the US?
Absolutely, like saying that a "computer-implemented invention must make a technical contribution in order to be patentable", then correctly deleting article 3 which said
and then introducing an amendment to recital 12 which says
The JURI amendments were full of that kind of silly tricks.
The average cost of a court case in the US to get an invalid patent declared invalid, lies between 1 and 2 million USD. Suppose that for some reason in Europe it would even only cost 1/4th of that. What will companies pick when they can choose between a licensing deal of EUR 50,000 or such a lawsuit? I really don't think it's a good argument to say "We must keep the law as unclear as possible and let case law figure everything out."
The directive is not worse for Free Software than it is for small companies. Neither have the means to obtain the defensive patent portfolio you need to survive in an economical system that allows software patents. Making an exception for Free or Open Source software could actually help monopolies. Microsoft would probably grasp this chance with both hands to crush anyone with an annoying patent they can't get a license for: just release a mediocre product based on that idea that's open source until the other person is squash