Appeals Court Makes It Easier To Dump Software Patents

An anonymous reader writes "While software patents are still legal, it appears that the Court of Appeals for the Federal Circuit, sometimes known as the nation's 'patent court' has just made a decision that will make it much easier to reject software patents for being mere 'mental processes'" rather than an actual invention. This could allow the Patent Office and the courts to reject many software patents."

Comments

Fixing the symptom

[afidel]

Isn't this trying to fix a broken process by fixing the symptom rather than the cause of the problem? I mean wasn't it just this week that we noticed that the last million patents were granted in 5 years versus the 80 it took for the first million? There's not that much more innovation going on today, we just have more patent abuse going on. Perhaps we need to have a higher fee for patents held by someone other than that original assignees, say $5k per year, this way small inventors don't get hosed, corporations are more willing to give up unneeded patents and will file fewer applications, and the patent office will have more funds to properly vet applications instead of throwing up their hands and rubber stamping everything and letting the courts sort it out.

Re:Fixing the symptom

[Mindcontrolled]

To be honest, doing application in traditional tech fields, mostly metallurgy, casting, automobile tech, most of the stuff I do in the US leaves me with the impression that the USPTO is going by a deny first - ask questions later strategy. You might want to find real points to criticize - perhaps start with the retarded common law system...

Re:Fixing the symptom

[gman003]

The process is broken because the two organizations are working at opposing ends. The USPTO is trying to clear out a (significant) backlog by approving pretty much everything, rationalizing that "the courts will clear out any bad patents". Meanwhile, the courts are thinking "you know, we really don't know much about science, we should probably step back and let the experts at the USPTO make the calls".

Naturally, that lead to a metric shitload of bad patents being granted and enforced. Which triggered far more patents to be filed for, making the whole thing pretty much pointless.

If the courts do start throwing out bad patents, though, the system starts working better. If it's a particularly moronic patent, it might even get thrown out during discovery, keeping court fees in the four or even three digit range. It won't make the system perfect, but it will make it less bad, which is a good thing.

Re:Fixing the symptom

[poetmatt]

However, guess when these denials don't happen first?

When it's expedited requests.

Re:Fixing the symptom

[Oxford_Comma_Lover]

It's not exactly "reprisals." It's more a question of job evaluation. An examiner who lets a bad patent through is doing a bad job. But an examiner who fails to let a good patent through is also doing a bad job.

Re:Fixing the symptom

[Mindcontrolled]

Haven't made enough of those to really give you an answer here - but our archive should give me a somewhat significant sample. I'll see if I can spot a pattern there, but honestly, I don't believe it. Personally, I get my stuff approved in Europe faster than in the US. Except for computer based stuff, which usually gets smacked hard in Europe - but such applications I usually only see coming from the US to be filed in Europe, so I can't really comment on the granting practices in the US in that field. I just object to the "USPTO is rubber-stamping everything"-meme. It's just not true.

Re:Fixing the symptom

[jellomizer]

There's not that much more innovation going on today [Citation Needed]

15 years ago when I was in college...
Classes were done with either Black/White Boards or Over Head projectors with a clear film and a Dry Erase marker.
They were computer generated projectors available but they were very expensive and not widely used.
In college I was one of the few students to get a Cell Phone. Most students didn't have one and they relied on a Phone Plugged into a phone system. My Moderns small form factor phone was about 1 inch thick, 2 inches wide, and 6 inches tall. This was a small phone. Black and white display, and it needed both an analog and digital cell receivers in it because digital voice transmission wasn't widely used. My Computer was one of the fastest PC I can build a Duel CPU 1 Core Pentium 200 chip with 64Megs of Ram, and a 6 Gig hard drive, and a SVGA card that can handle 1280x1024 resolution display with 32bit color. A campus of 6000 students shared a single T1 line and it was really fast. Rich students had a DVD player most didn't. And most music was still played on CD's and Downloading music was just barely getting popular, but only with the Techies.

Now compare that with today we are actually much further along then we realize.

Now compare USA in 1776 - 1856. Not that much improvements in technology, yes there were improvements but not as much in that 80 years as we would notice in the last 15 years.

There is a lot of in ovation that goes on today. And I see this as a good plan in the right direction, it isn't that we have to many patents, but a lot of them especially with software patents are too simple and any programmer could come up with the idea when they are approached with the same problem.

Re:Fixing the symptom

[BenoitRen]

To me all of that sounds like improvements in existing technology rather than innovation. CPUs got faster, mobile phones got smaller, etc.

Re:Fixing the symptom

[crutchy]

Re: "Perhaps we need to have a higher fee for patents held by someone other than that original assignees, say $5k per year, this way small inventors don't get hosed, corporations are more willing to give up unneeded patents and will file fewer applications"... you are not serious are you? The small inventors have enough trouble getting funding for real R&D and patent applications, and $5k/year is peanuts to those who are abusing the system (big corporate). Charge more and you'll lock small inventors out all together and patents will be the sole domain of the super-rich corporations. I think its a good first step for the courts to acknowledge that dodgy patents are dodgy because even if someone fixes up the uspto, so much damage has already been caused by the millions of (probably) dodgy patents already issued, so the only way they are going to be invalidated is in the courts.

Re:Fixing the symptom

[afidel]

Did you miss where I said for those who are not the assignees? If you are the inventor and you haven't sold your patent you pay the small fee applicable today, if you're a corporation holding a patent or a patent troll hoarding a bunch of patents in the hopes that you can submarine one then you pay the higher fee. Motorola Mobile has 17,000 patents, that would be $85M/year which would be a non-trivial expense and I would be willing to bet they'd give up a lot of them that weren't worth the $5k/year to have around.

Re:Fixing the symptom

[HiThere]

Maybe it's only rubber stamping things filed by US companies? Whatever the reason there are HUGE numbers of totally silly, obvious, or clearly invalid patents granted. And it costs millions of $ to fight each one through in court to invalidate it, *IF* you dare take the risk.

Re:Fixing the symptom

[Gadget_Guy]

To me all of that sounds like improvements in existing technology rather than innovation. CPUs got faster, mobile phones got smaller, etc.

And a car is just a wagon without a horse. It is amazing how simple and obvious everything is if you state it in simple and obvious terms.

Why is it that CPUs got faster? They didn't just deliberately make the chips slower 15 years ago so that they could sell you a faster version every year. No, there were physical limitations that needed to be worked around. New techniques needed to be invented to reduce the power requirements and crosstalk within the chips. New manufacturing processes and materials were required to be able to print the circuits at smaller and smaller sizes. These are the innovations. Faster CPUs are the results of those innovations.

Re:Fixing the symptom

[geekoid]

Or maybe you are wrong.

There are not "HUGE numbers of totally silly, obvious, or clearly invalid patents granted."

there are, however, HUGE numbers of totally silly, obvious, or clearly ignorant people on slashdot who wrap their whole patent opinion on head lines and sensational stories.

Re:Fixing the symptom

[geekoid]

sigh..

Innovation:
1. something new or different introduced: numerous innovations in the high-school curriculum.
2. the act of innovating; introduction of new things or methods.

innovate [in-uh-veyt] Show IPA verb, -vated, -vating.
verb (used without object)
1.to introduce something new; make changes in anything established.
verb (used with object)
2.to introduce (something new) for or as if for the first time: to innovate a computer operating system.
3.Archaic . to alter.

Why people on slashdot can't grok what innovate means is really beyond me.

I can slap a clock on shoes, and I have innovated.

Re:Fixing the symptom

[geekoid]

I like your idea; however if implemented, I would like to see the addition fees you pay, 7.5 yeas, and then again at 11.5 years, be reduced. right now they are about 2000 and 4000 respectively.

Re:Fixing the symptom

[sjames]

Now compare USA in 1776 - 1856. Not that much improvements in technology, yes there were improvements but not as much in that 80 years as we would notice in the last 15 years.

You mean other than the submarine, the steamship, the hot air balloon (human flight!), lithography, the battery, the cotton gin, powered spinning, and the Jacquard Loom. Gas lighting, the arc lamp, the tin can, the preserving jar, the steam locomotive, the miner's lamp, photography, the spectroscope portland cement, Braille, the typewriter, the sewing machine, the electric dynamo, the telegraph, vulcanized rubber, anesthesia, the smallpox vaccine, the facsimile machine (sort of a primitive fax that worked over telegraph lines), antiseptics (and with them, surgery that might not kill you), rayon, pasteurisation, and a bazillion other things.

Compare to faster computers and smaller (but less reliable) cellphones.

Re:Fixing the symptom

[Thing 1]

The annual price for registering the patent should double every year. Either the company is making enough money from it that they can afford the "tax", or they aren't, and they give the patent to the public domain so that everyone can benefit from the innovation.

I remember seeing a patent on "using a laser pointer as a cat entertainment device", and I think it's really funny that someone went to that length; with a low entry fee, we might see many silly patents -- however, they would likely only last a short time if the fees kept growing, and the idea was not something bankable.

Re:Fixing the symptom

[mysidia]

Perhaps we need to have a higher fee for patents held by someone other than that original assignees, say $5k per year,

How about... start it off at $2000 a year after the date of the original application for patent, and for any year that anyone other than the original assignee holds the patent have a per year scaled cost cost based on number of years since the patent application, to reflect the increased cost to society of tying up that past invention from use by new inventors. The invention must be really valuable if you're wanting to hold the patent for its entire duration, no? So... why not charge for that capacity, at increased cost the longer the patent is held EG

Year 0: $2000
Year 1: $4000
Year 2: $8000
Year 3: $16000
Year 4: $32000
Year 5: $64000
Year 6: $128,000
Year 7: $256,000
Year 8: $512,000
Year 9: $1,024,000
Year 10: $2,048,000
Year 11: $4,096,000
Year 12: $8,192,000
Year 13: $16,384,000
Year 14: $32,768,000
Year 15: $65,536,000
Year 16: $131,072,000
Year 17: $262,144,000
Year 18: $524,288,000
Year 19: $10,485,760,000
Year 20: $20,971,520,000
Year 21: $41,943,040,000
Year 22: $83,886,080,000
Year 23: $167,772,160,000
Year 24: $335,544,320,000
Year 25: $671,088,640,000
Year 26: $1,342,177,280,000
Year 27: $2,684,354,560,000
Year 28: $5,368,709,120,000
Year 29: $10,737,418,240,000
Year 30: $214,748,364,800,000
Year 31: $429,496,729,600,000
Year 32: $858,993,459,200,000
Year 33: $1,717,986,918,400,000
Year 34: $13,743,895,347,200,000

Re:Fixing the symptom

[crutchy]

a patent is a patent. if the original assignee sells a patent, then the new owner has paid for all the priveliges of the original assignee. should someone who builds a house pay less annual rates than someone who buys a house? while i can understand where you're coming from with your idea, it seems a bit descriminatory and anti-competitive, and if the costs are higher for someone who buys a patent, that buyer will either/both expect to pay slightly less for it in the first place because the cost to benefit ratio increases for the buyer and hence the value of a bought patent reduces compared to now where the value of a patent to the original assignee and someone who buys one is the same, and also any corporation who buys a patent under your scheme will of course merely pass the increased cost off to their customers anyway (probably with added margin for the trouble), and this certainly won't put a dent in the market share of patent trolls like Microsoft, so everyone is worse off except Microsoft.

Re:Fixing the symptom

[ppanon]

Sure, but many or most of those inventions you listed were made in Europe. The US was primarily an agrarian nation for most of the time range you quoted, whereas Europe had more urban areas, denser population which made communications easier/faster, and a class system that made it easier for the upper classes to study and research. Of course a meritocracy has important advantages over a class system, but these are less significant in an agrarian society, Having >80% of your population toiling on farms tends to take the wind out of the sails of innovation (apart for farming implements and techniques).

Re:Fixing the symptom

[testadicazzo]

Generally prefer incremental changes and improvements over major overhauls. In this case however I am inclined to agree with what I take as the spirit of your argument. Currently I see fixing the patent system as a real possibility, partly because many of the patents existing are so ridiculous and bizarre that they attract the attention and ire of even less technical and less informed members of the public, business, and government. If the system were to be so reformed that only sufficiently complex 'business methods' or software patents survive, this might mask the fundamental problems with patenting ideas, allowing it to survive to the profit of patent lawyers, but to the detriment of human innovation and small business.

Re:Fixing the symptom

[testadicazzo]

I would recommend the book "math you can't use" for a thorough and useful discussion of the problems with software patents. The FSF has many good things to say on the issue.

Re:Fixing the symptom

[sjames]

Some were invented in Europe, some in the U.S. However, they were nevertheless technological improvements actually seen and enjoyed in the U.S. In many cases, things actually invented elsewhere saw their first wide deployment in the U.S. with a great many incremental improvements in the process. Many of the incremental improvements we see today happen in Japan, Korea, and Taiwan.

The next 80 years saw a lot more invention in the U.S. It should be no surprise that the pace of invention was slow for a while, there was a significant upheaval involved in the Revolutionary War and in forming a new country after.

Re:Fixing the symptom

[poetmatt]

Wha? Bilski and other cases have resulted from a far too relaxed view on patentability. Clearly invalid, obvious patents have indeed been granted due to not being specific enough. That was clarified again by the federal circuit. Long story short, too much has been patented on things that are not patentable.

Re:Fixing the symptom

[Jose_X]

>> There are not "HUGE numbers of totally silly, obvious, or clearly invalid patents granted."

Before I consider how much lol to exert, I will need more information.

I am one who thinks that the inventiveness bar for being granted a US patent (35 USC 103),

NON-OBVIOUS ..

[kinda like that homework assignment in that tough class your taking this semester: it's really "non-obvious"]

to a person having ..

ORDINARY ..

[you know, like "average" and "unspectacular"]

skill in the art,

is

a) ever so slightly
b) somewhat
c) almost a foot
d) a foot and 8 inches
e) a meter
f) 2.84 meters
g) > 5 meters

higher than ankle height.

Re:Fixing the symptom

[ppanon]

The next 80 years saw a lot more invention in the U.S. It should be no surprise that the pace of invention was slow for a while, there was a significant upheaval involved in the Revolutionary War and in forming a new country after.

Sure, but a big part of that is that the industrial revolution finally made it across the ocean a few decades after it really got going in England and France. You need industrialization to provide force multiplication to the farmers, free up agrarian population for the growth of cities, and lead to more people with exposure to a basic technical education needed to push the state of the art. Again, there's a reason why most of those inventions you mentioned were made outside of the USA, and the ones that came from the USA were mostly from the tail end of that period. Apart for the steam locomotive and the steamship, most of the USA population wouldn't have known about any of those bazillion things you mentioned. Heck, antiseptics didn't really get widespread adoption in the USA until the Civil War's battlefield medicine acted as an indisputable proving ground. Also note here for a timeline of railroads in the USA, how the early ones came from England, and when people in the USA actually started inventing significant improvements.

There's a certain argument that, while slavery was the rallying point, the Civil War was also an economic war caused by the industrialized NorthEast wanting to keep its monopoly on industrial processing and preventing the previously agrarian South from becoming a competitor and cutting them out of the loop. In other words, the Revolution was possible because the USA had started to develop a localized industrial base, but the Civil War happened because that industrial base was no longer localized. That industrialized base is the incubator of innovation. So it should be no surprise that, after recovery from the Civil War, the rate of innovation as indicated by patent submissions would pick up.

Re:Fixing the symptom

[sjames]

Yes, in the next 80 year period, the U.S. transitioned to a strong industrial economy and started originating a lot more of the breakthroughs and innovations. We went from telegraph to telephone and radio. We got electric lighting and powered flight. The phonograph and moving pictures became common forms of entertainment. People traded horses for cars. The assembly line was invented. Essentially, the world transitioned to a form we would basically recognize today.

In spite of those massive changes, we did NOT produce as many patents as we have in the last 15 years.

Re:Fixing the symptom

[ILongForDarkness]

In my experience shitloads are always using the imperial measurement scale.

I agree with your post though, in general I don't think anyone wants to be the one to invalidate thousands of patents held by Apple, Microsoft, Oracle, IBM etc. Huge political fallout: "Oh you are ruining american competitivieness, etc etc". Imagine the dreamland that people wish to live in where you can export all the manufacturing to the developing countries and keep all the patents, trademarks etc in the US. Oh crap you mean all those hordes of engineers/manufacturers etc that we built up can now copy our programs? Oh crap.

Irrational ruling

[macraig]

Too bad the submitter didn't read the ARSTechnica article about the same ruling; it was a more impartial analysis and demonstrated how, even though the ruling appears to favor the ultimate abolition of software patents, it's such an illogical ruling that it probably won't really help to serve that purpose, other than perhaps persuading other courts to think more critically about software patents. "Unless it's too complicated for a human to do the math"? Good grief.

Re:Irrational ruling

[macraig]

Yeah, yeah, so TFS did link to the ARSTechnica article and I didn't notice it. My point, per my comment's title, is still the same; by itself, as a legal precedent for eliminating software patents, this ruling is useless because of the gaping irrational exception the ruling allowed. The summary ignored that aspect in favor of confirmation bias.

Re:Irrational ruling

[SLi]

It's nowhere near as confused as the ruling by a UK top court a few years ago about software patents.

A traditional argument for software patents has been that "look, it does have a physical effect - it makes electrons move in a certain way". The UK court ruled that algorithms are not patentable, unless the algorithm in question has an effect on the computer that makes it "a better computer", such as by "making more memory available for programs" or "making it run faster". Now I'm actually convinced that it wasn't a desperate decision by the court in order to keep software patents either: UK's response to an European Patent Office questionnaire about whether algorithms should be patentable largely repeated the conclusions of the court, and it's hard to say that it was a pro-SWPAT stance, although it certainly wasn't an anti-SWPAT stance. It was plainly and simply a confused stance which did not make any sense at all.

Re:Irrational ruling

[macraig]

It's sad that courts feel this need to qualify their rulings in such a ridiculous fashion, and ruin their value in the process. Do they fear the corporate and political party lynch mobs, or do they really lack confidence in the substance of the rulings themselves?

Re:Irrational ruling

[Theaetetus]

Too bad the submitter didn't read the ARSTechnica article about the same ruling; it was a more impartial analysis and demonstrated how, even though the ruling appears to favor the ultimate abolition of software patents, it's such an illogical ruling that it probably won't really help to serve that purpose, other than perhaps persuading other courts to think more critically about software patents. "Unless it's too complicated for a human to do the math"? Good grief.

It also directly goes against In re Beauregard.

Re:Irrational ruling

[rmstar]

It's sad that courts feel this need to qualify their rulings in such a ridiculous fashion, and ruin their value in the process. Do they fear the corporate and political party lynch mobs, or do they really lack confidence in the substance of the rulings themselves?

Don't be too sad. I think most judges do not really understand the issues, so are ruling from their interpretation of text of law and a more or less vague idea of what is going on. This isn't really their fault, as they are not programmers and software engineers. So I think the judgements are reasonably good. They are a precedent in the right direction.

Re:Irrational ruling

[StormReaver]

"Unless it's too complicated for a human to do the math"? Good grief.

Yep, we're back to square one. All that will happen now is that patent applications will include bloated calculations to pass the "too complicated for a human" math test. They won't actually be necessary for anything except printing more money...err...getting the math patent approved.

It's good to see that our judiciary is stuffed with people who utterly fail at basic literacy.

Re:Irrational ruling

[martyros]

Yep, we're back to square one. All that will happen now is that patent applications will include bloated calculations to pass the "too complicated for a human" math test. They won't actually be necessary for anything except printing more money...err...getting the math patent approved.

But surely that would make an easy defense against infringement? "Ladies and gentelmen the jury, consider the math in the equation in the patent: [Wild gibberish including lots of greek symbols, imaginary numbers, and infinite series]. Now here is the math in our algorithm: y=2x+1."

Re:Or not

[IDK]

If you print out the database, you can still do it by pen and paper?

Bad Patents

[hhawk]

Anything that helps reduce the # of bad patents helps here.

The Bottomline is any start up in the tech sector is going to face so much pressure from patent holders it clearly had a chilling effect on real innovation.

Patent Court

[sir_eccles]

Having an actual patent court staffed by judges who are dedicated to patent law and nothing else and cutting out juries (who don't know anything about patent law) like they do in the UK would be a great step forward.

Re:Patent Court

[WebManWalking]

As for cutting out juries, no can do, at least not in the US. From http://www.house.gov/house/Constitution/Amend.html here's the 7th amendment, in its entirety:

• In Suits at common law, where the value in controversy shall exceed twenty dollars, the right of trial by jury shall be preserved, and no fact tried by a jury, shall be otherwise re-examined in any Court of the United States, than according to the rules of the common law.

Re:Patent Court

[Ecuador]

You are suggesting that we can't do something in the US due to the Constitution?
What you are saying would be very funny if it was not dreadfully tragic...

Re:Patent Court

[king neckbeard]

I'm not so sure about that. Specialists in patent law tend to be pro-patent, likes GIles Sutherland Rich, who is the one that ruled on In Re Allapat and State Street. The best track record of a recent high judge on such matters seems to me to be John Paul Stevens, whose early background was antitrust, a field that tends to not be fond of legal monopolies like patents and copyright.

Re:Patent Court

[JustinRLynn]

It's more like we can't not do something, that is to say, it is directly required.

Re:Patent Court

[MartinSchou]

Are those 20 dollars adjusted by inflation? Because 20 dollars would buy a hell of a lot more stuff in in 1776 than it would in 2010.

The relative worth of $20.00 from 1776 is
$515.00 using the Consumer Price Index
$10,500.00 using the unskilled wage index

Re:Patent Court

[shentino]

What would help is not letting lawyers give wiseass patent experts the boot during voir dire.

It seems like lawyers know damn well enough about their own bullshit that they WANT juries to be dumb and impressionable.

Re:Or not

[wierd_w]

Actually, you can pen&paper logical operations like that as well, using logical notation.

This sort of appeal could cripple the likes of MPEG-LA and pals though, because codecs are essentially fancy math at the core.

Hell, is you accept mathematical notations, then hand-written code samples with worked out code execution should also be applicable, because at the most abstracted level they are essentially tha same kind of construct.

(Mathematical notation is just a limited language for dealing with quantities, using a linear process. Logical notation does the same thing but with relations, and computer source code does both and has extentions for repeated processes.)

Re:Or not

[ailnlv]

formally speaking, anything that a coputer does can be done by hand with pencil and paper

Credible threats

[sjbe]

Isn't this trying to fix a broken process by fixing the symptom rather than the cause of the problem?

Yes, though I would argue that any improvement is a good thing. This potentially significantly lowers the standard of proof needed to get a patent thrown out which means they are then less likely to be used as a threat. A patent that is not a credible threat is much less worrisome. It doesn't solve the legal cost issue but it might very well keep some legal threats from being made in the first place.

Missing the point.

[tgd]

The problem isn't software patents, the problem is lousy patents. There are just as many lousy business practice, gene and physical patents as there are lousy software patents, and there are perfectly valid ones of each type.

(And, yes, I know that /.'s knee jerk reaction to software patents is a pathological "patents bad!", but there are vanishingly few people on here who seem to understand how patents actually work, how to read them, and how to understand what is covered.)

All invention is a mental process. The point of patents is to protect the people who expended the effort to DO that mental process, and to reward them in exchange for documenting that process for others. Most ideas are obvious is retrospect to people who are experts in a field, but its a fallacy to believe it was obvious to those same people BEFORE they saw it. That's the reward for a good patent -- being the one who actually saw it and let other people know.

Re:Missing the point.

[plankrwf]

--The problem isn't software patents, the problem is lousy patents.
--There are just as many lousy business practice, gene and physical patents as there are lousy software patents,
--and there are perfectly valid ones of each type.

Pitty slashdot is not Wikipedia, or else I would have thrown in {citation needed} at that last line.

Kind regards,

Roel

Re:Missing the point.

[vux984]

I see Invention is 2 separate steps:

1) Identifying a need
2) Answering that need with an invention.

A patent should be a non-trivial / non-obvious response for step TWO.

What I see is a lot of people conflating the difficulty answering the need vs having the spark of genius to identify it in the first place.

A lot of patents are really obvious answers to questions that no one had thought to ask yet.

The little tripod/table thing to keep cheese getting stuck to the lid of a pizza box is an obvious solution to the problem it solves.

Ask any qualified engineer to actually solve that problem, and sticking a little plastic "load bearing tripod or scaffold" on the center of the pizza is a no brainer.

So how did we go for decades without them?

Because it took a moment of "genius" to actually ask the question.

But the patent system isn't supposed to reward people for asking questions... its supposed to reward people for coming up with innovative answers.

As you said: "Most ideas are obvious is retrospect to people who are experts in a field, but its a fallacy to believe it was obvious to those same people BEFORE they saw it."

If the idea for a solution is obvious once the problem is posed its not supposed to be patentable.

That's the problem with patents. If I read an abstract for "a method to suggest movies based on viewing habits of customers on a computer"

Then I'm going to envision a database of movies, and who watched them, and some algorithm for correlating a given customers viewing habits with others similar, and then suggest movies that other people like him watched that he hasn't watched yet. Maybe i'll even take into account how they rated the movies. So there will be a feedback mechanism...

The exact algorithm used to suggest movies might be validly subject to a patent, but not its general structure.

Thus other vendors should be free to implement their own "movie suggesting systems using databases and correlating viewing habits on a computer" without any fear of interference from this patent. Indeed the patent is too broad.

If it instead was a patent on a distance algorithm for relating viewers relative to movies viewed that was actually innovative such that an expert in the field wouldn't suggest its general form without even seeing the answer, that might be a valid patent.

My point is being the first person to think of doing
X shouldn't be rewarded if the method of doing X is obvious to anyone now that you've thought to ask them too.

Re:Missing the point.

[JimFive]

All invention is a mental process.

No it's not. Edison didn't just think, "Hey, I could put a filament in a jar and run electricity through it to make it glow" and patent that. He actually did the work to make a light bulb, and it took a lot of trial and error. It was not just a mental process, but a physical process.

Most ideas are obvious is retrospect to people who are experts in a field, but its a fallacy to believe it was obvious to those same people BEFORE they saw it.

The idea isn't at issue, the implementation is. Obvious, as related to patents, should mean that, given a statement of a problem, an expert in the field would come up with a substantially similar solution. While it is true that figuring out which problem to solve can be the hardest part of invention that is outside the scope of the patent process.
--
JimFive

Re:Missing the point.

[Chris Burke]

All invention is a mental process.

All invention(verb) is a mental process. Not all inventions(noun) are mental processes.

This is the very similar to the difference between something that can be described by math (the parabolic path of a thrown baseball), and things that are math (the actual math of the parabola, as expressed in a symbolic language).

Software is math. Software is a mental process. Software patents are patents on math. Software patents are in fact the problem.

Lousy non-software patents are also a problem, but a different one.

Re:Missing the point.

[russotto]

but there are vanishingly few people on here who seem to understand how patents actually work, how to read them, and how to understand what is covered.

No, it's the pro-patent side who misunderstands how patents actually work. They think that just because a claim can be interpreted narrowly, that it will be interpreted narrowly. In fact claims are interpreted broadly, and prior art narrowly.

Most ideas are obvious is retrospect to people who are experts in a field, but its a fallacy to believe it was obvious to those same people BEFORE they saw it.

Except, of course, when they were.

Re:Missing the point.

[WorBlux]

Being the first to market is a pretty big reward in itself, as is being the foremost expert on a technology. There is also the extra cost of a monopoly put on consumers, and cost of filing and litigation. All in all, it's not clear that patents create value, even if you limit the analysis to the better 20% of current patents.

Re:Missing the point.

[MrMatto]

... Edison didn't just think, "Hey, I could put a filament in a jar and run electricity through it to make it glow" and patent that. He actually did the work to make a light bulb, and it took a lot of trial and error. It was not just a mental process, but a physical process.

Edison did not come up with that idea. He stole most of his "inventions" from other people and claimed them for himself. http://www.enchantedlearning.com/inventors/edison/lightbulb.shtml

Re:Does that qualifies?

[elsurexiste]

...it is a plain good engineering in response to market's demands. Not something I would call an invention or an innovation.

I would say the opposite, actually :) .

If Then Mental Processes

[retroworks]

If I patented "if then" logic lines, then I could sue the court for using my patented logic either way. Unless they generated randomized rulings. Which sometimes does appear to be the case. If on the other hand, I was issued a patent for a process of applying for patents, then I'd be protected from any other patent holder. The court needs to think very hard about applying patents to mental processes.

How do we mod...

[ilsaloving]

How do we mod up the Court of Appeals (+1 insightful)?

Not convinced

[uigrad_2000]

If you said "4 colors is not always enough to color a map", but refused to give an example that took more than 4, then I would not be convinced.

If you say that not all software patents are bad, just most, but fail to give an example, then I'm not convinced either.

Re:Not convinced

[martyros]

If you say that not all software patents are bad, just most, but fail to give an example, then I'm not convinced either.

The original mp3 audio compression technique represented a non-negligible investment of R&D time and dollars to invent. But once known, it was easy to copy.

The whole point of the patent system is to allow a companies that invest in R&D to be able to recoup that investment.

Re:Or not

[killkillkill]

Or done by hand with A Bunch of Rocks

Re:Or not

[ArhcAngel]

silly...a coputer is simply a dual core puter.

A better idea...

[mark-t]

... instead of making it easier to dump them, how about just making it mandatory?

Software patents are an abomination, and should be eliminated wholesale... they are tantamount to patenting thought, however impractical it might be perceived to be to do such computations manually.

So, base logic is safe for now ?

[unity100]

my thoughts are still my own ?

Got it backward.

[AnotherBlackHat]

Rather than test for what can and can't be patented, I'd rather have a test for something that can't be subject to a patent.
Suppose, for example, I could say "My software runs on hardware that existed before your patent was filed, and therefore my software doesn't violate your patent.".
Then it would be possible for me to write software and be relatively certain it didn't violate any patents.

See, I don't want to invalidate your stupid patent, I want to avoid being sued.
I only decide to invalidate it after you sue me.
Which means I have to read and understand your stupid patent.
Which means you have a strong incentive to make your stupid patent as difficult to understand and as confusing as possible.

Even a very restrictive affirmative test is better than a lax negative one.

Re:Does that qualifies?

[ColdWetDog]

Classic Apple: take something, remove stuff from it (in this case, half of the thing), innovation!

This is an old technique. In motorcycle racing we used to call it "adding lightness".

Re:Or not

[ynp7]

I have a math coputer at home. It helps my puter pute floating point operations.

But what about integrated circuits?

[BlueCoder]

What they do is logical in the same way. A person can mentally walk the steps to evaluate signals going into and out of digital and analog circuits. Does this mean we actually need to apply copyright law to circuits as I suspect? What about the past 100 years of electronics cases using patent law, I doubt the courts would be so willing to go so far even if it is the logical conclusion.

An circuit is something physical but what it does is a mental process although it does it very fast. But if tangibility is the definition of what can be patented I have to wonder if a specific book title would not also qualify as a patentable object.

Re:But what about integrated circuits?

[WorBlux]

I believe circuit layouts weren't protected for a long time but now use a grant similar to copyright. Chip designers would embed trademarks into the silicon, so that any 1:1 copy could be trademark infringement, and patent various aspects of operation (still do for the most part).The main difference here though is that although circuits can be described mathematically, they aren't actually math.

Re:Does that qualifies?

[fyngyrz]

This is an old technique. In motorcycle racing we used to call it "adding lightness".

funny... that's what I call it when the motorcycle rider falls off...

Never happen

[Weaselmancer]

Hurrah for the great idea, but anything that shuts down software patents would cripple the economy.

Here, read this.

Google gets hit with a few patent lawsuits over Android, so it responds with the $12.5B buyout of Motorola Mobility. And everyone knows why they did it. Patents. They have to make a proactive move to protect Android. So, how much of that value is in their patent portfolio? Now imagine that value suddenly going *poof* in a puff of logic. Now multiply that across every software company you know.

The effect would be devastating.

I wish it were possible, but I just don't think anything is going to come along and just suddenly devalue software patents. Too much value is tied up there to simply vanish.

Re:Never happen

[geekoid]

Yes, it might lower their stock value, but it will lower all of there stock value; assuming the market gives a crap.

Having an inflated value vanish is normal economics.

Re:Never happen

[js_sebastian]

Yes, it might lower their stock value, but it will lower all of there stock value; assuming the market gives a crap.

Having an inflated value vanish is normal economics.

Not necessarily: abolishing software patents will reduce the value of some companies (that have a good warchest of such patents and a history of knowing how to use them to print money or quash competition) and increase the value of others (that do not have quite as good a warchest, or even if they do are vulnerable to patent trolls).

Also, it would accelerate the growth of new, innovative companies, which would lead to overall economic growth.

Re:Never happen

[Weaselmancer]

I agree with you. In the long term abolishing software patents would stimulate economic growth. But you're going to have one hell of a bumpy ride getting there. In the short term it would be pretty disastrous.

I'd like to see software patents go away. But just know that there will be a cost. It'll be a real nightmare at first. A lot of techies are going to suddenly be unemployed.

Re:Does that qualifies?

[geekoid]

Then maybe you should get a dictionary:
innovate [in-uh-veyt] Show IPA verb, -vated, -vating.
verb (used without object)
1.to introduce something new; make changes in anything established.
verb (used with object)
2.to introduce (something new) for or as if for the first time: to innovate a computer operating system.
3.Archaic . to alter.

Re:Or not

[halowolf]

I had a Comptometer (http://en.wikipedia.org/wiki/Comptometer) when growing up. Hours of endless fun.

Bigger than that

[Weaselmancer]

Not on this scale. This would make the housing bubble look like a hiccup. Each company maintains a "war chest" of abusive patents. Motorola Mobile just got bought out at 63% higher than market value. at a sale price of $12.5 billion. And almost entirely for their patent portfolio to keep Android alive and viable. We all know that's the reason, even if Google is hedging the purchase in marketspeak.

Now imagine you're a shareholder.

Suddenly, the company you've invested in jumps for 63% over market value due to their patent portfolio. Then along comes a new ruling and it is essentially worthless. Up 63%, then suddenly zero.

See the problem yet?

If not, multiply that across the board. Imagine every single software company you know going through the same. Microsoft, IBM, Oracle, HP, Google...imagine all of them going through this. Their market caps being suddenly reduced by 20% to 80% or so. The war chest is suddenly useless.

Think of what that would do to the economy. Think of what that would do to your career.

Gives me the screaming heebie jeebies. I freaking *hate* software patents, but now that they are part of the calculation of the worth of your company - beware! If they go away we're all in some seriously deep shit.

Re:Bigger than that

[Agent0013]

Each company maintains a "war chest" of abusive patents. Motorola Mobile just got bought out at 63% higher than market value. at a sale price of $12.5 billion.

Suddenly, the company you've invested in jumps for 63% over market value due to their patent portfolio. Then along comes a new ruling and it is essentially worthless. Up 63%, then suddenly zero.

See the problem yet?

I don't understand why you don't see the problem with buying imaginary property for 12.5 billion. If you were a share holder of a company, would you rather they spend 12.5 billion on buying another company out so they can fight a money wasting lawsuit, or would you rather they innovate, create, and make money rather than send money to lawers. If all the companies lost their software patents, whatever drop in share price that occured would be even accross the board, unless the company is only a patent troll, they would stand to loose more. Plus the lawers will loose out. And since laws are made be lawers, I don't see them hurting their own cache cow and changing the broken patent system.

Such a pity that Apple have found a new way

[advocate_one]

to cause massive chaos using the European Community Design patents... which are subject to no review at all when being granted, merely have the right forms been filled in in the correct manner... the referenced Community Design being the one responsible for Samsung being blocked from selling their Galaxy items in Germany as they purportedly look like a thing thing with rounded corners that Apple have registered a community design for...

100 million+ lines of source code

[Jose_X]

Ask a patent examiner next time you think you've spotted one online (beware, they are usually friendly and considerate) how many of the hundreds of public open source lines of code they have carefully reviewed as part of their preparation to judging novelty and obviousness in the patents they go through daily.

Do not be shocked if the person replies 0.

Do not be shocked if the next "software patent" they pass overlaps with the state of the art.

However, should they get lucky, do not be upset at them if the next patent they grant is merely "non-obvious" to a person having "ordinary" skill in the art. The law requires them to grant that patent, even if it was TOTALLY OBVIOUS to hundreds of thousands of software developers who are ABOVE ORDINARY in their skill in the art (35 USC 103).

The US's pathetically low inventiveness bar goes without much thought when only a tiny fraction of the population is inventing and trying to raise hundreds of millions of dollars to take these inventions to market. But it comes with huge opportunity costs when you have when you have millions of software developers impacted. A software development, manufacturing, and distribution network is but a mouse click away for each development iteration.

Many think the entire patent system (which, btw, does NOT recognize independent invention, something that appears to be unconstitutional whenever speech like software is involved) should be scrapped. At least for process patents (like software and business methods utilizing no more than widely accessible materials), there should be near universal agreement they aught to go.

Re:Or not

[ObsessiveMathsFreak]

Anti-aliasing is done with math, and could be done by hand, but it would take forever (from the ars article).

This goes to the heart of software patents. They are not patents on computer software. They are patents on mathematical algorithms which need not be run on computers at all.

They can be run on water powered wooden clockwork, by pen and paper, and on Turing machines built in Minecraft. They can be run by choreographed groups of Hungarian folk dancers. They can be run in the human mind.

And an exclusive licences to run these algorithms, without even having to write them, are being handed out like confetti by an office that's not even supposed to be granting them.

Re:Or not

[hvm2hvm]

Depends on the size of the paper and whether you have properly positioned people ready to read signs written on it from a distance and relay them forward.

Most are stupid.

[StillNeedMoreCoffee]

The idea that throwing out a "mental process" is an absurd metric. All software is the codification of a mental process and some definitions of an Algorithm are that they are so clearly defined that you can implement them with a pencil and paper. So those are bad rulers to measure the fitness of a software patent.

The problem is to find how high the bar should be set for "mental processes" to be unique and new enough to have patent protection.

Because of the speed of innovation, and such quick obsolecence, and not wanting to bring the system down in a screeching halt, maybe the best idea is to move software (and firmware and hardware if they are just hardcoded logic like software) into a special category and have the patent last for say 5 years max. Then you can get your money out but not retire on the residuals, you'll have to re-invent something.

The length of patents (and copyrights) is too long to foster innovation and leveraging ideas in this computer age, The "community" approach to this broken situation has been the Open Source movement, where we have said, we will all work to provide something free for all of us. Breaking the owners and royalties forever model.

Or maybe a comprimse. Give the patent but let its value (percentage) depreciate over some period so you still have benefit but diminishing over time.

(let me think, maybe I should patent or copyright that idea)