Slashdot Mirror
EU to Redefine Scope of Software Patents
karvind writes "According to story on ZDNet, the European Parliament (EP) has enlisted the help of intellectual property lawyers to amend the directive on the patentability of computer-implemented inventions so that companies are prevented from patenting pure software. According to article: "The ongoing argument over patents in the software industry revolves around the distinction between physical inventions that use software -- such as a car braking system -- and pure software." (See also this earlier story about the EU and software patents.)"
Patent lawyers get to write the EU patent laws. I can hardly wait. *Groan*
Stop the bullshit. Software should not be patentable, not pure software, not embedded software, NO software. Patent lawyers have been preaching for years that software is already patentable, you just have to word the patent application right: "A device which uses the following algorithm to..."
If this can be implemented so as to disallow pure SW patents, it will probably be good enough to bring some real sanity back into the patent business. Especially when the WTO has to decided between US-spawned stupidity (one-click) and what the EU doesn't allow.
Make sure everyone's vote counts: Verified Voting
I have no idea how this patent mess will turn out, but it's none the less a Good Thing that they are looking into the issues possibly ahead.
Keeping my fingers crossed...
.: Max Romantschuk
Pure Software seems to be a very very strict definition. Would it catch things like JPG compression which isn't software? After all GIF issues were a royal pain. What about one-click? That could be a business process. Are those patentable in the EU?
I can see many patents getting allowed as weasel speak can turn something into "hardware" or similar quite easily.
* Winners compare their achievements to their goals, losers compare theirs to that of others.
Well, the software patenting lobby never uses the phrase "software patents". No, they merely want "computer implemented inventions" to have the same protections as other inventions in the industry. Most inventions produced by industry today have software in it. Like car braking systems. If they can't patent software, they can't innovate in (for example) car braking systems, they say.
The solution of course is to allow patents on the car breaking system. If something novel is happening there (outside of the software) then it should be allowed. It shouldn't matter whether the system is controlled by software or a black box full of small midgets.
Another solution: Allow the patents, but make it absolutely clear that no patent can be infringed on by writing, publishing, downloading or using software on a normal computer.
)9TSS
The European Parliement (elected by Europeans to function with EU level competence). Has effectively given the bird, to the blatantly pro-Patent the undeomcratic (European Council, various heads of state waffeling about crap). The EP, emasculated the original Patent draft, only to have the Council, make an ever *more* pro Patent document.. bang it's chest and say to the Parliment "Go on... change the text if you have the balls". Turns out the Parliment did have the balls to change the text. Better still, this directive would *eliminate* patenting pure software... so, to the earlier poster who questioned "jpeg" the answer is no... under this proposal, patenting jpeg is out, patenting mp3 is out, patenting the double click, one click shopping or XML is similarly out. Patenting a software controlled system that influences the physical world, like an alram panel, or a mechanism to open/close a garage door with software somehow *is* allowed. Personally, I welcome our new Democratically elected, Representative, Neo Natal Federalist government. But I'm not paying any taxes !!!
OTOH not all IP lawyers are bad, so lets hope the ones being used by the EP genuinely want to do the right thing here.
It is highly unrealistict to expect Patent offices to have enough expertise not to pass stupid patents. As it goes, even if patent is invalid, nobody wants to challenge it as it would bankrupt them in court before they win.
So why not just set expiration of patent to 6 months, with possibility of prolongation for another year if nobody put the objection in the first 6 months ?
Such a legislation might actually have a chance to benefith both the inventor (ok, its employer realistically) and community (as after max. year and a half it will be free to everyone.)
Besides, I'm not sure that it's even true. Many of the major technological breakthroughs of the past 150 have never been patented. It would never have occured to Crick and Watson to patent their knowledge of the structure of DNA, or for Einstein to patent the creation of nuclear power through mass-energy equivalence.
And true, there have been many tech breakthroughs driven by patents in the last century, but are any of them more staggering than the isolation of new elements, the discovery of Periodic Table, the theory of radiation or the structure of the nucleus, or the germ theory of disease, electromagnetism, or Archimedes' Screw, or the wheel, or... Well, you get the idea.
Would the chemical battery had developed as fast as it did if Volta had patented the technology and tried to charge Davy, Faraday, Carlisle et al, instead of publishing in a journal for the general good of mankind? Would that have encouraged scientific progress, or discouraged it?
[Clue : it's the latter]
It may seem like technology is growing exponentially, but it seems to me the number of distinct "Eureka" moments in the 20th don't massively outnumber those of the 18th.
If the rate of progress has increased, it's because technology begets technology : microscopy => structure of the atom => electon microscopy, and because political, social and economic reforms have opened up science to a class beyond the professional dilletante.
Athletic Scholarships to universities make as much sense as academic scholarships to sports teams.
Patent lawyers are no legal experts, they are technicians.
-- The Online Photo Editor - http://www.phixr.com
Can the USPTO even go back from here? What about the tens of thousands of software patents that have been issued already? If software patents were ever reformed, would they be 'grandfathered in' until they expire? I mean people/scum (depending on the situation) may have spent considerable time and money securing each software patent they obtained. What a mess.
This is Slashdot.
If you make a joke, however obvious it maybe that it is a joke, you must say that it is a joke. For example could surround it with tages like , or (if it is at the expense of a group you end it with the word "ducks".
The preferred approach used to be to make a joke that follows one of the acceptable Slashdot templates such as "in Soviet Russia...". However as these have fallen out of use and Slashdot's users have expanded outward and downward even these may be moderated troll.
In short if it is not clearly marked as a joke, how do you expect moderators to recognise it as such? For example, if it was TV there would be some canned laughter to indicate when you should laugh therefore you should use a suitable substitute here.
What do you expect? Intelligence? Literacy?
Do you also realise that by suggesting that not everything was invented in the US in the last few decades, you have probably ensured that a significant number of Slashdotters will have decided you are anti-American, if a few of them have mod points you will get modded down again.
A patent is there to protect an inventor so that he might profit from his intellectual efforts, or in many cases his, results of a great deal of time invested in stumblinbg accrss something through "dumb luck" which is more likely a finally honed sense of recognition built upon said investment coming to play.
Now we get to the gist.
Software as implementation of an algorithm should not be patentable. Said algorithm should only be patentable if it results in a new and novel process or product BUT if it's a process patent, the process must be related to a product AND a product must be more than the look and feel of a presentation.
Thus, a new compression algorithm would be patentable but the presentation of the uncompressed data in a window on a computer screen would not be. The 'one-click' order does not depend upon a novel algorithm so it falls into the presentation category and would not be patentable, although a particularly efficient 'one-click' algorithm might be. Treating a specific sort of information as an 'object' is not novel because many sorts of information might be treated as objects. The FFT is a good example of an algorithm which would deserve patent protection.
Separating algorithms from implementation schemes
makes software patents a lot easier to enforce because the algorithm reduces to a mathmatical formula.
The Sarnoff-Armstrong patent war over phase vs frequency modulation would not have happened as one patent would have been a derivitive of the other. (this problem is not new)
Nope, you're not. You're saying that the innovation progress in the software field doesn't benefit from software patents. The patent system is built to encourage innovation - it's its purpose. If the legislating party (congress, parliament, bundestag, riksdag, whatever) deems that software patents indeed are hindering innovation in an area, it's perfectly legible to NOT award patents to the innovations in that area.
I have a really elegant proof for Fermat's last theorem. If this sig was only a bit longer...
Is the problem with the patents themselves, or merely with the way that they are awarded?
The problem with patents is in two areas:
A company can use patents to stifle competition in a particular market, while not actively exploiting that market for their own gain. Perhaps they already have made a considerable investment in one technology, and don't want to throw away that investment due to a new superior and cheaper technology. Instead of making any investment, they just sit there and do nothing, but just harass everyone else (Much like the telephone companies vs. community wireless).
There are also other companies who just buy up patents from bankrupt startups and use them to blackmail other smaller companies for profit.
For a large company with a patent portfolio and lawyers, they can either counter-sue or arrange a cross-license. But for smaller companies, who can do neither, they just end up having to pay up.
Vintage computer adverts: http://www.vintageadbrowser.com/computers-and-software-ads
In fact the US got a leg up in the 19th century by blanently "borrowing" the ideas for many industrial items from England, completely ignoring the Crown's patents. I fully expect this situation to be repeated this century with the China and India "borrowing" from the US and EU.
On the contrary, Microsoft is running to catch up with patent free open source. Internet Explorer is NOT leading the way, it is trying to catch up with Firefox. Patent free Firefox is the innovator.
The real patent problem is that if you could patent office and stop anyone else from producing office productivity tools for the next twenty years, then there is absolutely NO REASON to innovate. You might as well sit back and watch as people are FORCED to buy the only software available.
The problem with patents is that any areas they're applied to will tend to gravitate towards a natural lifecycle identical to the patent lifetime (17 years, iirc).
In industries such as, say, grand piano making, where the natural product lifecycle is very long, patents give an effectively zero window of opportunity - they just don't last long enough to be worth getting. It could be argued that in this case patents don't put enough stress on the industry.
In industries such as bioresearch, where creating new products is very costly, the tendency is to just produce knock-offs. As a result the natural product lifespan is probably in the region of 30 years or so (after that the drugs will cease to be effective or will have been surpassed by academically-researched alternatives). In these industries patents give a medium-length (relatively speaking) window of opportunity for getting ahead of one's competitors, thus encouraging innovation. The stress that this puts on the industry speeds the average product lifecycle up to approach the lifetime of the patent.
In the software industry, the average product lifecycle lasts for somewhere between 1 and 7 years - call it 3 years. The lifetime of a software patent is almost 6 times that. In such a fast-moving industry, this is an effectively infinite window of opportunity, resulting in companies theoretically being able to get a big heap of patents then sit back and relax (or fail to get patents then die). The stress that patents put on the software industry is thus misdirected towards lengthening rather than shortening the product lifecycle, and just results in lots of protesters rather than any actual progress.
I would accept software patents if they were of duration = 2 or 3 years, but then they'd be so short that, given the speed of the patent system and legal system, they wouldn't be worth filing for. Anything longer than that and they can only be harmful, with progress proceeding despite patents rather than because of them.
Shouts to Clausewitz for the "stress" metaphor.
For the love of God, please learn to spell "ridiculous"!!!
Actually, the idea behind the "forces of nature" wording is that you should be able to patent a physical device that includes a software component, but not the software itself (that's covered by copyright), nor the algorithms you used, nor the idea of controlling the device by software. So you would not be barred from patenting your computer-controlled camera that follows motion just because it contains a software component, but you would not be able to patent the software or the algorithms themselves, nor would the patent cover all possible computer-controlled cameras that follow motion, but only those who take advantage of the "inventive step" you made.
This approach has the advantage of outlawing software patents while not rejecting the directive outright. This is good because a) if the directive is rejected, a new one will appear shortly, b) a positive but restrictive directive would reinforce the ban of pure software patents, c) lawmakers seem to feel that their job is sanctioning new laws, not avoiding them, so they are less likely to reject the directive altogether.
The problem with this approach is that it is confusing to many people (/. readers routinely discuss the subject, yet there still are differences in interpretation). This confusion is an invitation to abuse. Current European patent legislation bans software patents altogether, yet the EPO ignores the rules and grants them all the same. Nothing indicates that the EPO will honor the "forces of nature" wording, even if it's imposed on it.
In all, while the "forces of nature" approach may be a workable solution, it's not worth it: all this confusing concession buys us is the supposed bonanza of innovation that patentability would fuel. However, it has been shown that patentability not merely in the software but in the whole computing arena actually stifles innovation, rather than encouraging it. There again, it may be all that we manage to bargain from this situation.
Consider the following: no proof has ever been established to show that patents makes more innovation possible. Some research suggest the opposite might be true.
In software the opposite has been proved true. Microsoft is the perfect example - they have for the last 10 years often said that they would invest more in research, and yet what has happened is that almost all of the extra research investment is going to lawyers.
Also, if the true culprit was the patent system - why let patents cover MORE things if the system is as clearly inadequate as most people can agree it is.
I considered the problem of accepting software patents and ensuring the patents where in fact proper patents (this part is important, since if this is not done patents ARE a hindrance to innovation and conducting business - and also one of the main arguments for the pro-sw-patent lobby that the european system is not like the american system because we check the patents more thoroughly - this however is sadly less and less true).
To do this my unscientific conclusion was that it would demand so many ressources that the system would only be open to the extremely big corporations because getting patens would be very prohibitively expensive (or the state would pay thru the nose to pay for the patent administration and thus give higher taxes).
I'm not really convinced for the case of patents - but if patents incorporating software was to be allowed by law (as opposed to the current practice in the EU of approving SW patents which by the letter of the law shouldn't be patentable), then at the very least the following 2 things should be applicable:
- Copyright should not apply for the original patented implementation of said invention
- Sourcecode for a working implementation should be supplied and available when locating the patent application
This, at the very least would restrict the patentowner from exerting his patent on related inventions and would also limit the patent to the exact prototype showed - whereas today software patents cover ALL possible implementations of the invention.The most oft used analogies are the car motor analogy where we have gas and diesel engines and one manufacturer can have a patent on the gas engines and enother on diesel engines - where as in software the patent would cover "engines" as a whole, including but not limited to rocket engines, jet plane engines etc.
Nice summary of the problem with patent durations. As you say, they cannot be suited to all fields; the duration is too short for slow moving fields, yet too fast for rapidly advancing fields.
;)
So let me throw in a wild idea that just occurred to me (if it occurred to someone else previously, you should have patented it
The duration of a patent needs to vary according to the field it's awarded in. Short for computing and biotech, but long for piano-making. How to decide the different durations? Setting them artifically seems like it's going to be a really difficult issue. So how about turning the problem back upon itself.
Rapidly moving fields will have loads of patents filed in them, whereas slower developing fields will just have a trickle of patents.
So why not tie the duration of a patent to the number or rate of patents being awarded in that field? Create something like the stock market that's self regulating - as more people buy, the price goes up.
That way you automatically shorten the duration of patents in fast moving fields, while not adversely affecting developments in piano-building say.
Plus, it has the advantage that it may serve to deter the filing of frivolous patents, as anyone who did so would be shortening the duration of any other patents they held in that field. You could argue that this could be done deliberately, to shorten competitors' patents, but the company doing so would still need to come up with an invention of their own.
I'm sure that there are plenty of practical difficulties, such as determining the number of patents in a particular field, and what field a new patent applies to, but hey, it's just an idea.
Most anti-patent groups accept that the former should be patentable
Excuse me? If a group is SUPPORTING the patentability of certain things, how the hell are they "anti-patent"?
Labeling people opposed to software patents as "anti-patent" is a flat out lie, malicious propaganda. A straw man attack. It's sad to see otherwise good and unbiased reportering falling for it and with their own words unwittingly slandering one side of the debate.
-
- - You can't take something off the Internet! That's like trying to take pee out of a swimming pool.
I'd rather they went a step further and prevented companies from patenting all software, pure or impure.
But since I don't expect they'll ever do that, how about a law that states "software that runs on hardware built prior to the patent application is not subject to the patent."?
-- Should you believe authority without question?