Username or EmailPassword
Your post is a very good starting point for discussion about patent reform, Thom.
One of my "pet peeves" is patents about "appearance".
For example, patents on "rounded corners" of a window on a screen, or rounded corners on a device. Patents for a particular "finish" on a product - even patents for *colours*!
I believe that all patents for "appearance" should be declared invalid, and that patents should only be granted for novel *functionality*. This move in itself would cut out a lot of patent-trolling. Edited 2012-07-05 23:25 UTC
"One of my "pet peeves" is patents about "appearance"."
You're right about this because because patents about appearance are redundant concepts. Trademark law more than adequately addresses this issue in a far more sane, understandable manner.
It's called Trade Dress.
Some products such as tablets are so minimal and basic that it's pretty hard to not look at least something like other tablets. To go as far as Apple wants to push the issue would actually cripple all competing products. That's not right. The Slate/Tablet form factor has been around for much longer than the iPad.
In such cases the packaging and logos are pretty much the only obvious differentiators.
Think about soda/soft drink cans/bottles. Many bottlers use the exact same containers from often times the same suppliers. The nature of this product dictates that the label is the only obvious differentiator.
Many years ago (and for a relatively short time) I was a Patent Examiner in Australia. From my recollection shapes WERE NOT patentable. Australia has/had a designs Office to register particular designs but even there I don't believe you could register an oblong, not even one with a concentric internal oblong - representing the shape of an iPhone/iPad/what ever else.
I think that government down-sizing in all/most jurisdictions has forced Patent Offices to reduce staff levels to the point where now they can no longer do a thorough examination so they grant a patent application and let holders of "prior art" (similar previously granted patents) sue the new patent holder. In my day this was to be avoided.
Basically the article here is very good but alas I feel without more pressure on governments nothing will happen. Legislators need to be squeezed where it hurts, then they will respond.
Seems like a simple solution would be to quickly reject the stuff that is obviously ridiculous. Mega corps ask for patents in droves; if it is something they really care about, couldn't they just resubmit?
EDIT: shamefully I know very little, but I assume there is a minimial price to ask for a patent. As long as this is in place and not prohibitive for the average joe, then this idea seems like it would be viable. The alternative would be a "patent submission tax" for companies over a certain size. However, if there's one thing that seems to be true in the US, it is that large corps seem to get more government benefits than small corps (and not just through loop holes - also through the result of lobbying). Edited 2012-07-06 10:24 UTC
Wasn't it in Australia that after a lax patent reform, somebody successfully patented the wheel? Edited 2012-07-06 11:11 UTC
I'm with you all the way on this. Just the other day I was helping a friend with his iPhone 3GS, and I set it down next to my phone (HTC Arrive). Twice I almost picked up the wrong phone when just glancing down; I had to actively look at them to tell the difference from three feet away. Thinking back to other touchscreen phones I've owned, I realized just how much they all look alike. After all, how many variations on the design can there really be?
Apple patenting the rectangular glass-and-metal touchscreen candybar phone form factor is just crazy. It was around before the iPhone (LG Prada comes to mind) and was an inevitable design choice given the direction all the manufacturers were headed.
I think most people have a serious misunderstanding when it comes to patents. They believe patents are created to ensure inventors obtain profit from their invention. The is a consequence, NOT the purpose, of patents. The focus on the rights of patent holders at the expense of innovation in society is the problem. As you mention, patents are to foster benefit for society at large. When patents serve to block progress, then something is amiss. Every patents should be examined for how it benefits society.
From this, it follows that all the inane patents on software and business practices SERVE NO USEFUL purpose to society at large. I like your list of changes, as they would swing the balance back towards benefiting everyone, not just the inventor.
I take issue with you on this point. As a former Patent Examiner I can tell you the purpose of a grant of a patent is to protect the intellectual property of the inventor. It is to stop others stealing/modifying original inventions so as to deprive the original inventor of his rights.
Of course, this does not mean I believe there is nothing wrong with the existing system.
That is where I disagree and I disagree based on the history of patents. It is very sad that, you , as a patent examiner don't seem to know why patents were created.
The purpose of the patent is disclosure of technique or technology under the promise that only the inventor will get to use it for a limited time. When you say it's to protect IP of the inventor, you are missing the point of the whole idea of patents and their historical significance.
I think the misunderstanding is between the original purpose of Patents (as stated elsewhere), and the actual practice of patents today. I would agree with your assessment of how patents are viewed and used today. Obviously you would know! The arguments is whether or not the current practice matches the orginal intent.
fretinator said: I think the misunderstanding is between the original purpose of Patents (as stated elsewhere), and the actual practice of patents today ... The arguments is whether or not the current practice matches the orginal intent.
As I've said several times, I don't believe the Patent Offices of the world have the resources to keep up with the demand for their services. They seem to have adopted an approach of "let it through and the 'prior art' holders will sue if they feel aggrieved enough".
When I worked for the Australian Patent Office this was most certainly NOT the case but times have changed.
Governments either need to cough-up with the resources or simply get out of the way.
I agree with the general sentiments of this post - it is just ridiculous the way things are today. A company with big pockets (like Apple) can stifle competition by relying on patents (like for a single search function or the shape of a mobile phone) that should NEVER have been issued.
Peter Edited 2012-07-10 05:23 UTC
I think you, JAlexoid, and fretinator are absolutely right. The justification for granting patent monopolies has shifted. Initially, the whole reason the public government allowed private monopolies was to provide an incentive to disclose trade secrets. In other words, the patent system was envisioned as a public benefit rather than a private entitlement. Today patent holders are typically considered the main beneficiaries, and any shred of public benefit is completely sidelined and obfuscated. Patent holding companies feel entitled to own patent monopolies regardless of whether the system produces a public benefit or not.
As for the lack of resources, I think there's a combination of factors at work. I've already mentioned some recently here:
The number of patents is always increasing (close to 200k granted per year in the USA alone, no clue on foreign patents), the resources that count as prior art (traditional book/magazines & new online media & existing products) just continue accumulating as well. So regardless of what cost cutting measures the patent office might try to implement, the per-patent costs (after inflation) should logically be projected to rise indefinitely for a constant level of service.
Since the PTO can't/won't raise their prices indefinitely, the level of quality simply has to drop. We could allocate more resources to help the PTO cope for the time being, but in the long term it cannot scale along a field as large as IT without impeding it. Edited 2012-07-10 06:23 UTC
The number of competitors is irrelevant.
The people who working with the same problems are using the same methods. Because experienced people have a common background / common sense.
Even the lightbulb and radio has track conflicting claims.
Now the situation is completely absurd.
I agree; problem is, as you say, the legislation is decades old and has not been brought up-to-date to mirror the world today.
In my day (early 70's) most of the Apple "look and feel" type patents would not have been given Patent protection. Maybe they could have been a registered design but I feel they are too broad even for that.
As I've said before, the Patent Offices of the world simply don't have the resources to keep up any more.
Do you have a working legal definition of a "software" patent? What exists in the US are process/method patents implemented using a computer. Some of those methods could be implemented in hardware and hardware functions can be translated and implemented in software. Where do you draw the line and what working legal definition do you use? I know it when I see it?
I hate the patent mess as much as anyone, but reform is a sticker problem than some hand waving can accomplish.
Let's pretend that I find a way to compress video to 1/10 the output size of any other codec, while retaining all existing quality. This is a novel and useful invention, but if there is no software patent, then I cannot protect my methodologies. While I agree that most software patents are bad, there needs to be a way to protect a novel algorithm. Although the math itself cannot be protected, there needs to be a method to protect novel means of deriving a solution.
then I cannot protect my methodologies
And why should you do it?
Just write the codecs/apps and sell it.
Do not disclose "methodologies" if you fear clones.
If it so trivial that even children could reimplement it,
then your methodologies worth nothing. Edited 2012-07-06 18:19 UTC
That's part of the problem right there; comparing software to pharmaceuticals. Software, which has a very high rate of evolution and very low costs to develop and implement, if it deserves patent protection at all, should not have the same term of protection as pharmaceuticals with a comparatively low rate of evolution and very high costs to develop and implement.
Show me several years of development, testing and FDA approval prior to general use of bounce-back affect at the end of a scrolling screen or similar bullshit patents and it'll maybe make a little more sense to talk about the half-decade protection rackets being granted on mathematics and business processes. Edited 2012-07-10 13:00 UTC
Bill Shooter of Bul,
"No, that's still math."
Agree math algorithms shouldn't be patentable. But we've seen idiotic patents over software which are not necessarily considered math, like XML case against microsoft. Software shouldn't be patentable whether it's considered math or not.
I like the definition posted earlier: any software running on commodity hardware should be in safe harbour from all patent infringement claims.
"But then again, the prices we pay aren't that bad, considering everything its capable of."
Well, I'd be curious to see an exact number of how much overhead patents are responsible for. But there are other factors that we shouldn't ignore like the monopolization of features, and technologically inefficient engineering workarounds that ultimately reduce product utility for consumers. Even corporate consolidation driven by patent acquisition has eliminated products from the market. It's difficult to put a dollar value on these kind of social costs.
Regarding the weakness you point out, that seems to be a direct consequence of assuming that a human being must own the rights to a patent, not an organisation.
But while I know treating corporations as people isn't approved of, this is one place where it does make sense. If a patent is created by a team being payed to do so by their employer, it's not unreasonable that the employer be the owner of the work they produce. You won't achieve your goal of encouraging innovation if the people paying for it are afraid that the work will simply walk out the door once done.
I believe the simple solution to this problem is either a patent is filed as personal or as contract labor. In the case of a personal patent, then the authors scheme would stand. In the case of a contract labor patent then the patent would be filed as belong to a corporation. In this case, the patent would remain valid for the term and the only provided mechanism for transferal would be in the event that the owning corp. was purchased in its entirety.
If the company is bought by another company in part or in full then the original entity no longer exists and the patents enter the public domain.
If the company splits then the original entity no longer exists and the patents enter the public domain for all to use. Maybe the patent portfolio is split between relevant inventor departments during the process of divy-ing up resources. The requirement is that the patents remain with the originating entity as long as it exists; they are not transferable assets to be shuffled about on the chess board.
No more buying and hording of patent war chests. The amount of resources pissed away in this childish game
The approach that immediately leapt to my mind for working around this issue is to alter the clauses covering restraint of trade and inventions in the employment contracts.
Many currently state that any products or inventions created under the employ of the company become property of said company, but this would become invalid under the proposed system.
It's also not uncommon for companies (in my country at least) to have a clause preventing you from working for a direct competitor for a certain time period after leaving. If sections such as these were amended, then the result could be something like:
Any products or inventions developed under employ of the company must be licensed exclusively to the company for $timeframe. On discontinuation of employment, all products or inventions developed under employment of the company must be licensed exclusively to the company for a period of $timeframe2 or until the expiration of the patent. Whichever is shorter.
I got an idea, create a different website to put all those opinions and stop contamination osnews.com of your personal agenda againts patents.
Is tiresome, if this site could provide some kind of filter, but no, we are forced to view your personal attacks and view point, witch really I don't care any more, it was relevant before but now is tiresome.
For the record, I think the patents system is broken, but not to the level of being obssesed day and night. Edited 2012-07-05 23:54 UTC
You are forced to read OSNews?
Not really, but is not fun anymore, I've been reading this site since 2002, yeah 10 yeears, and is not the site I got addicted to read daily, now is more your personal tool, you don't care about the reader interests like before, you care about your own interests.
So, yeah, I'm not obligated and I prolly will stop visiting this place because now is boring.
Btw, I'm sure I'm one of the few people who never used adblock in this site.
Are you forced to write OSNews? Delegate. If no one wants to take the mantle, admit defeat and close the site. It's also true that your personal blog is pretty sparse for the last few years. Why not write your opinion pieces there and link to your blog as a one of the "less important" side bar stories?
I dunno Thom, the site is less about OSNews, more about OSrude. Rude people, bigots, zealots, no one listens to anyone.
However it seems that it is consistently the same people always complaining about the content of the site. If they don't like it, why don't they go somewhere else? Or even create their own?
They keep coming back and consistently complaining about the same things. Why do they keep coming back? There must be some content on the site that isn't available elsewhere, or maybe it is just their routine?
I personally come back because I like some of the content. But I do find myself biting my lip most of the time. I used to be more vocal, but now I'd rather not get in to a circular argument.
I fall in to this mould:
* I like Apple
* I like Microsoft, especially the development tools
* I enjoy programming
* Former BeOS user, mainly on PowerPC (I owned a BeBox)
* Mainly use Mac's at home. My choice. It works for me.
* Mainly use PC's at work - mostly Windows.
* Have built/configured Gentoo from a bootstrap image
* Have owned Nokia and Sharp Linux based hand helds
* Was a Palm user and developer
* Have no particular use for Linux in my daily life, outside of appliances.
* Like iOS, and own devices that use it
* Like Windows Phone 7.5
* Have coded extensively for Windows mobile and various mobile LINUX devices
* Have not used Android since 1.5, probably would like to own an Android device 4.1 now.
* Owned Amigas
* Owned Acorn Archimedes and programmed demos on it
* Have used NextStep and OpenStep and have coded for them both.
* Have owned a newton
* Own a classic Mac (both 68000 and PowerPC based)
* Have written device drivers and low level code
* Have compiled an entire desktop class operating system from source that was *not* LINUX or LINUX based.
* Have written commercial code for embedded devices, handhelds and dashboard computers (Taxi/Goods/Transport) ans a hella lot of Windows and Mac
You can't nail me down or call me a fanboy, I'm not. I have a varied experience and I'm not specific at all. But at the moment, engaging in any meaningful conversations on OSRude is almost impossible because of the attitude of the minority.
Agreed!! I'm starting to sympathiyse with that guy a fews years back that ran osviews.com for a while.
Too much venting opinion by a minority (Thom mostly) and not enough actual OS news.
This is Thom's personal blog effectively. Sometimes there is interesting OS news, but mostly (very) biased anti-Apple stuff.
But if Thom were to quit the site would die. The other OSNews team members might as well get lost in a magical forrest and we probably wouldn't notice until Christmas.
So, I guess we better be happy with the few gems we find burried in all the other stuff.
It would have helped if Thom was somewhat technical savvy. Maybe if they can add such a person to the team and somehow keep him active it would increase the number of truly interesting articles and outside the mainstream news. Someone that is open minded, not extremely biased and has seen tecnology change over the years.
Not getting in to rude arguments with his/her readers would help too!
Wouldn't it be much easier just to ignore those people?
I can understand you'd like to strike back when people attack you, certainly if it's unfounded, but it doesn't help and just makes things worse.
If you really want to reply in such a case do it politely. It makes you look more intelligent. When you explain the how and the why people can take it or leave it, but if you're rude or insulting you just provoke more nasty replies.
Now I know I haven't always been overly nice to you, but I do appriciate the time and energy you put in and know a lot of people fail to understand this.
To have an extra active person on board would help, the site and you personally.
I am a bit shocked about personal threaths against you and your family. Time for an aangifte.
At home, which I'm very far away from right now, I have this Seiko watch.
It's pretty rare and can be "docked" on a small keyboard and act as a screen to run BASIC programs on. I have two keyboards, a normal one and one with an embeded printer.
The problem is I can't connect the watch to either. The one with the printer keeps printing "connection error".
There are no visible connectors. I do as the manual says, but never does it even seem to connect.
Do you have any clues?
Agreed. This site has turned into a software patent echo chamber.
Conversely, it doesn't exactly make it true if dissenting opinion is drowned out and actively characterized as trolling.
Besides that being wrong, it has absolutely nothing to do with my comment. Stop posting.
I am glad you are creating conversation about patent reform. Hopefully continued conversation can get tje snowball rolling.
There are however serious weaknesses with many of your proposals. I will just point out one. To reduce the holding term to 5 years would be a disaster. If anything they should extend the length. This allows costs to be amortized over longer periods. If an invention is very expensive to create and the window of profitability too short, there will be no incentive to develop the invention, thus defeating the whole purpose of the system.
I do agree with you that there needs to be some mechanism to ensure market availability of inventions. One of the worst practices of companies (especially fossil fuel industry) is to gain ownership of technologies and squirrel them away so no one can utilize them. We can't prevent the resell of patents, but we could require fair market value access.
I truely don't understand why you get voted down. Your post is not rude and merely expresses a different view.
Why do readers here downvote a reaction that is not in line with their belief?
I think part of the blame lies with the current moderation scheme. I've felt for a while that the instant gratification of seeing a post's score drop after being modded down is an incentive to abuse the system. I think if they were to change it so that, say, five downvotes for the same reason (troll or spam) are required to cause the score to decrement by one point, it would eliminate a lot of insightful posts being hidden by fanboy modding. The system should still hide the downvoted comment from the voter, since if you're downvoting them you obviously don't want to see the comment anymore.
Make that the same for up-votes? Horde groupthink up-voting of a bit silly posts is also too often visible... (also as in: vote-whoring posts that don't really say anything, but will make enough people nod in agreement "hey, that's how I think!")
Sure, it's not really so distorting, but it's a bit sad when you see quick "clever" almost-one-liners that were up-voted through the roof simply for targeting groupthink, while some long & interesting posts down the line barely can get to 3-5.
But I guess that's also the typical dynamics of comments, people jumping quickly on first posts, imposing such pace, and then it quickly dies out.
"But I guess that's also the typical dynamics of comments, people jumping quickly on first posts, imposing such pace, and then it quickly dies out."
Actually, this is probably a consequence of the voting system used on OSNews, which artificially restricts voting later in the discussion. Unless one abstains from voting (and posting) at the beginning of the discussion, they'll loose their ability to vote on posts later in the discussion when only those with duplicate accounts or who didn't post can vote. I prefer a strait forward voting system, because the current one can't be considered representative.
It's more than just that. For example, look at the most recent main page discussion: http://www.osnews.com/comments/26165
What does the first post even say, is there any real substance to it? I mean, it's "positive" of course, easy to agree with - but why is it at 9, the highest-rated post of the discussion? You don't have to vote.
Furthermore, note that a post fairly down the line ( http://www.osnews.com/thread?526160 ) is also quite high ...while, apparently, what it says and its source are not even factually correct - but it's something many would wish for, and overall it's still closer in nature to a one-liner that few earlier longish, interesting, and lower-rated posts.
No, it's also us, it seems that twitter-like messages are just more "effective" on people; probably we're just much less likely to read and/or appreciate the long ones - which required some thought to write them down, and require it when reading. Plus, considering how the short posts are less meaningful, less specific - they have that vote-whoring effect, more likely to be found agreeable by like-minded group.
Not artificially (? ...are there any natural moderation systems? ) restricting voting later in the discussion would probably still promote mostly-one-liners, possibly even more so.
Maybe, one day, a post-webcam technology can determine if you actually read posts before moderating, maybe coupled with separate "baskets" of votes (per post length category), and perhaps even neural monitoring to determine the effort both when writing and reading
I wonder how that would work out...
Limit the PTO to granting five patents a day.
This will result is only a few, but very strong, patents being granted. Of course there will be a giant fight to be one of the five and big corporations will win most of the time, but they already win most of the time.
For the rest of us life becomes much easier. Out of the five per day less than 100 a year will be computer related. Now it becomes possible to track those few patents and simply avoid the technology and associated lawsuits.
It should also ensure that standards don't contain patented technology since there is no way someone with one of these strong patents is going to share it.
I think some of your ideas are actually good, as in I agree with them - but many of them would never have any chance of being adopted (in the states anyway) along with your 2nd idea - abolishment of software patents.
There are many vocal people in the US pushing for patent reform , but most of them are being deceitful. It is not because they want the system fixed - it is because they want to avoid any possible outcome that might result in the elimination of software/method/design patents. There are too many people with a vested interest in these kinds of patents around - they are desperate to keep this particular reform off the table. They talk about trying everything but that.
The real solution, that solves about 95% of the problem, is simple: just abolish these three types of patents (really two, as software patents ARE method patents). They never should have been allowed in the first place and both are arguably in direct contradiction of the stated goals of the system.
As you said, software is adequately protected by copyright, the the notion of design patents (when fairly applied) could be covered under trademark law with a bit of massaging. The dirty truth is that most people for "patent reform" in the US are against this solution...
We need to stop beating around the bush - many of those other ideas you had would be great and all, but the only one that really matters is abolishment of types of patents that don't make any damn sense. Without abolishing the "bad" patent types, nothing will really get solved...
General utility patents, pharma patents, etc. might have problems - but they are no where near as serious. Edited 2012-07-06 01:12 UTC
Headshot... Double-Kill... Multi-Kill... Ultra-Kill... M-M-M-Monster-Kill...
"Patent applications must be accompanied by a working prototype"
That would prevent me from seeking funding for the prototype. Requiring a working prototype at the time the patent is submitted only benefits wealthy corporations.
I should be able to obtain a contingent patent relatively quickly, and once I am able to build the prototype, I would submit it to get the patent validated. Of course there should be a time limit between the two events to prevent me from sitting on the patent.
I agree. Also what if your invention is as big as a house? How you gonna carry that to the patent office? Walking down the street one day and you come across the space shuttle replacement being towed to the patent office for validation. Not very sensible really!
There is no need to radically change the entire US patent system - just the system for software patents.
The US mandatory patent pool for aviation works very well.
Many patents are non-trivial and very complex. Pharmceuticals, for example, are typically extremely slow (~14 years) and extremely expensive ($500m+) to bring to market. That is why these products need such a long patent protection process.
I like your list, Thom.
I too dislike the use of big corporations handing their patent portfolio to their team of lawyers and having them destroy the little guy.
I like the idea of revamping the patent system to be a system that rewards the innovator, big or small.
Regarding your employee/employer patent ownership question, wouldn't it be the company? I don't know if it's common practice in The Netherlands, but in the States with some companies, you, the employee, sign a legally-binding document that states any ideas you come up with are owned by the Company, and you don't have rights to patent and make money off of them, even after leaving the company.
You are being compensated by the company for your efforts. If you're someone who does come up with patentable ideas, then perhaps you should leave a company that would do this and start your own business and build up your own protections against others who would steal your ideas.
The goal should be proper rights to earn money by being the first one to discover a brand new idea and encouragement of small business creation.
I'm not sure I understand your "McDonalds" scenario.
If you do get an idea for a new device and don't want to give it up to the company, you can setup a Trust and assign the idea to the trust as a separate legal entity from yourself. That way, the trust is what owns the idea and then patents the device, and not you who signed the legally-binding agreement.
I ran into a patent attorney who said that there were ways around such existing common legal agreements.
Patents are SUPPOSED to be useful to someone skilled in the area the patent covers. However, virtually all patents today are run through a team of lawyers who convert the patent into legalese in an effort to broaden the patent, and to obfuscate the workings of the patent to prevent people skilled in the area from actually using the patent. This is COMPLETELY UNACCEPTABLE!
My suggestion: No lawyers can be involved in the writing of a patent. Period. All filed patents are run past someone skilled in the field, and if they can't understand the patent in its entirety, it's rejected out of hand.
Thom, you must add JLF65's suggestion to your list!
Patents must be readable by skilled engineers, and not only that but skilled engineers should be able to reproduce the entire invention from the information filed in the patent (or publicly available information referenced directly by the patent).
Another point that you don't mention is how do you handle the case where the invention was developed independently? I don't have an answer, just bringing up the question.
One last thing to point out is that today some applicants intentionally enumerate all possible workarounds for an invention so that competitors will have trouble engineering alternate solutions around patents. Again, I don't have an answer for it.
I'd make it that any patent owned by an organisation is non-transferable. Therefore if a company gets bought, their patents are expired (ie: Google buying Motorola, everyone buying Kodak)
However if a person owns the patent they should be allowed to transfer it to whomever or whatever they want but the timeframe till it expires should continue from the point the patent was originally granted.
This would allow an individual to come up with an idea, get the patent then make some money out of it from a quick moving business (or other individual) that would be able to monetise it. However as soon as a patent ends up in a non-individual's hands, it can not be transferred from that: only expire.
[What if a business puts its patents in the name of one of its people? Well, then they run the risk of that person leaving with 'their' portfolio of patents and the company has no recourse]
I just think there has to be room to allow creative individuals to profit while preventing the distortions of reality that happen when corporations patent everything that isn't tied down. Edited 2012-07-06 02:02 UTC
I had a different idea to address much of the same points. We have 2 problems: how to get a patent and what is the effect of the patent. So
1) Patents must be granted based on publications in government-approved professional peer reviewed journals. This has a lot better chances of weeding out bogus/vague/trivial submissions than anything PTO officials can ever dream of.
2) A patent entitles its holder to reduced tax rate on income from the invention. No monopoly rights of any kind.
Second part is probably too radical to be implemented. Alternatively some mix of the following is at least an improvement on the current state:
* Limit patent term to 5 years, at least for "technology" patents.
* Force FRAND-style licensing for all patents.
* No enforcement by non-practising entities. Aggressor must prove practising for the whole duration of his damages claim.
* Patent holder must actively defend it to preserve his right for future litigation, a-la trademark law. This prevents someone from hiding patents and then going after successful implementations when they generate enough income to become an attractive target. Edited 2012-07-06 02:17 UTC
Even if they were settled the original filings would still be public. Do you have any evidence that this actually happens? I haven't heard this from any trustworthy source, but I regularly hear otherwise, e.g. PG quote above.
Many of your ideas (untransferable ownership of patents by their inventors) won't ever be peacefully accepted. Equally well, or maybe even easier, you could call for abolishing the whole patent system.
I you want a better analysis of the current situation watch the excellent "patent pollution problem" talk: www.youtube.com/watch?v=nfH8iyNjpYo
For me, the best solutions would be to either: abolish the patent system completely (it has always been restraining innovation, never promoting it) or go back to the origin: grant patents in exchange for trade secrets that cannot be reproduced by other experts in the field in a week of work. Only then, the society (ultimately patents are granted by consumers) gets a net benefit from the deal.
I think software patents can be done in the spirit of a patent, properly. This will also require copyright reform, though.
So, copyright law currently applies to both the binary and the source.
First, change copyright to 5 years.
Then, give the option of using a software patent INSTEAD of copyright. If you do this, the binary and source are protected as if it were copyrighted for 7 years. However, the source is immediately publicly available for review, and after 7 years, is released to the public domain. Now, grant this indiscriminately.
Congratulations, you've now just given an incentive for everyone to open source their stuff.
and set an maximum license fee for all patents. (At least do these things for software patents.)
BTW, copyright is different than patents. In a software patent, it's not the code being patented, it's the process that is implemented by the code. That process could be implemented by lots of different code (each of which is protected by copyright), but if all those different code implement the same process then they are all copying whoever invented/created the process in the first place, and that inventor/creator deserves some reward for the work he/she put in in creating/inventing said process.
Let's consider the notion that someone can spend hundreds of millions of dollars in R&D inventing a process, then implements the process in software; then someone else comes along and copies the same process spending minimal (or zero) resources in R&D, yet can get away with providing the original creators zero recompense because they implemented it with different code base. That is absurd and is a slap in the face of the original creator. It says, "Your inspiration has ZERO value, only your grunt work coding does." Sorry, I don't buy it. The fact is, inventing the process is the HARD part. Once you've created the process, the coding is just grunt work. Yet Thom and those in his school of thought would hold that the grunt work deserves protection (in the form of copyright) but the much more important R&D that invented the process deserves zero protection. It's backward thinking.
The "software is math" proposition is a philisophical outlook. I don't happen to agree with it (I don't agree that something is "math" just because it can be represented in binary form; nor do I necessarily agree that "math" is not patentable. I think such arguments, if used in the patent debate, must be proven or at least demonstrated, rather than merely asserted as if they are axioms. Edited 2012-07-06 03:39 UTC
I think that the assertion "Software is math" is a bit misleading. A piece of software can be described with mathematics, just like you could describe a piston with mathematics (dimensions, physical composition/atomic makeup, etc).
However, this doesn't mean that the problem the software (or physical device) solves isn't solved in an ingenius way! So, I feel there is room for software patents, as it seems they are there to encourage people to come up with inventions to solve problems.
That isn't to say the system isn't broken:
Slide to unlock? That's been featured on bathroom doors since forever!
First, a remark: What Apple used to fight the Samsung Galaxy design was not a patent but a design claim similar to what furniture and fashion makers use to protect against copies. So this has nothing to do with patents.
I think a better solution than all the proposed is that an inventor (person or corporation) should be required to document the expense they had in developing the invention. This would then be used in determining a maximum license cost and the period of protection. The documentation can be challenged in court in the same way that the originality of a patent can. Corporations should have stronger requirements for documentation than individuals.
This way, pharmaceutical patents that are based on years of developments and clinical trials can have long protection and high license costs, but a patent for, say, a magnetic power plug would have shorter protection and lower license costs.
"Patents should be non-transferable" - I don't agree here. Given all other proposals, e.g. short ttl, need for productizing etc., there's no harm in being able to transfer a patent, since that would hinder an inventor that only invents for the sake of it, not produces (yes, he could licence the patent etc., but that would be a hassle; speaking of licences: I think the European compulsory fair licences are good, and should be included in the US system).
No need imho for a "backup" inventor either, if the inventor dies, the patent passes to whoever is the heir. And as for employers of companies: a legal entity may hold a patent (as suggested by other commenters), so that includes both humans and companies (besides, most inventions aren't done 19th century style "weird scientist" Tesla kind of way, but in teams).
"In case of emergency, the state should be able to either temporarily or permanently nullify certain patents." - I don't agree here either. The state should be able to force licencing, even against a 0 fee, and if possible reemburse the patent holder. But nullify an otherwise valid patent is imho equal to theft.
"Patent applications must be accompanied by a working prototype" - I agree, to an extent. This is largely to prevent patenting of mere ideas. If one has thought about a certain method to do something, it must be tested whether even possible (impossible stuff should not be abel to be patented, akin to perpetuum mobiles), and in order to do that, a prototype is needed. The extent I'm talking about is that I'm not sure it's possible by a patent office, without enlisting an army of technicians and scientists, to verify whether a prototype actually implements the patent.
Above I said I didn't agree that software is "math", but didn't expound on that. I will do so very briefly here.
I look at Microsoft Excel, and in my opinion:
Microsoft Excel is neither a mathematical formula that was derived, nor a mathematical principle that was discovered. It's a state machine implemented as software, and that software was constructed, not derived or discovered like a mathematical formula or mathematical principle. This is why I don't consider software to be "math".
OK, that's my brief expounding. Now for some philisophical and metaphysical rambling.
I don't think the "software is math" argument is relevant to patents. Look at Amazon's "1-click" patent. Most people hold that that patent is absurd. But not because it's "math". The idea of "1-click" (including a description of how it works (in the case of "1-click" the idea and the description of how it works are essentially one and the same)) isn't "math", even if the software (i.e. the source code or compiled binary code) is considered such. And it's the idea of "1-click", not the software, that was patented. Or does the "software is math" crowd argue that the ideas themselves are "math" even before they are implemented as software (in which case, again the "software is math" argument is rendered irrelevant)?
I get the feeling that the "software is math" crowd feels that all mathematical principles already exist in the abstract, just waiting to be discovered, and therefore since they already exist, they can't be patented; and/or that all mathematical formulas already exist in the abstract, just waiting to be derived, and therefore since they already exist, they can't be patented. And they want to say that software is "math" (I guess because software can be represented as a series of 1's and 0's or can be interpreted using an abstract Turing machine, or some similar notion), and therefore all software already exists in the abstract, just waiting to be derived or discovered, therefore can't be patented.
If that's the line of thinking, then could one not take the next step and argue that non-software inventions also already exist in the abstract (see Plato or Aristotle), just waiting to be discovered, therefore nothing (software and non-software) can be patented? Or could one not at least argue that since the *idea* of every non-software invention can be represented as a series of 1's and 0's (ASCII text, for example), that therefore every *idea* itself is "math", and therefore every idea already exists in the abstract, just waiting to be derived or discovered, and therefore no inventions can be patented, software or non-software?
(Reminds me of a "proof" I once saw that said that since every idea can be generally described in 500 words, and every word can be represented as 1's and 0's, that therefore there is an absolute limit to the number of ideas that can and will ever exist. :p
I propose the following:
* a tax of amount T is required for granting a patent
* a "liberation" price is calculated as a function of the tax payed, f(T), and assigned to the patent
* if any one entity or consortium of companies, people or nations pays the liberation price to the patent holder, he is considered fully indemnified and the patent is then nullified. The technology is now in the public domain.
* other elements may be used in calculating the liberation price.
a pharmaceutical company invest 100$ millions in medicinal research, the output is patented paying a tax of another 10 millions dollar: the liberation price is set to 10$ millions times 100 = 1 billion.
The resulting medicine cures cancer, so other pharmaceutical companies, states, charities and sick people all pitch in and pay the 1 billion. Great profit was made by the company, but cancer cure is available cheaply and quickly.
A guy has bright idea about some useful gadget related to cellular phones, while he is sitting on the toilet. He gets a patent with 1000$ in patent tax. A single company quickly buys the patent for exclusive use at 10.000 and uses it in production. After the first market success the invention becomes a must, other phone companies decide that it is worth to pay 100.000 to a competitor and liberate the patent, rather than license it individually.
A bright idea is rewarded, but does not become a burden on consumers of manufactured goods.
- Exit patent, enter brevet. This covers the procedure. If you have to show a working prototype, you are less prone to horse around
-No brevets on mathematical software, mathematical principles, scientific theories and genes.
-I agree on shifting the burden of proof to the appellant, that should make them think twice before starting something.
- Inventor that leaves the company is still owner of the brevet that he presented in his free time and loses the brevet he created as employe of the company ( we suppose the company paid him to find exactly this)
- The peculiar case of the "invention by luck" is covered as follows: Brevet stays in the company, inventor receives a 25% royalty lifetime non transferrable.
This just leaves us with one big question. What do we do with the patents that exists NOW?
This is the real elephant in the room becuase WiFi data transmission ( just to make an example) has dozens of patents that covers each a slight detail of the system. So who gets it? Any idea?
Plenty of good ideas, some that need work, some that overreach.
I get the general impression that your anger about the software patents situation has made you go harder on hardware patents than needs be.
One example; if you abolish software patents, severely limit patent's lifespans (though it needs to be a lot more discretionary than five years. Many inventions won't have paid for themselves during that time), and make the bar far higher for getting patents in the first place, you really don't need to make patents the untranferable property of one or two individuals.
What drugs company is going to pump millions of dollars of research into developing a miracle cure only for one of the three dozen people who worked on it to get all rights to it?
I mostly like your suggestions with one big exception:
* Patents will no longer be assumed to be valid by courts.
WTF?! That's plain stupid. If you pay big bucks to file a patent application or whatever and it's granted after long examination, of course it should be assumed valid by courts.
If the patent isn't assumed to be valid, how should the aggressor argue that no prior art exists? Arguing something doesn't exist is basically impossible.
Perhaps it would be better to allow the other party to show that the aggressor's patent is invalid (as they only have to show an example of prior art). Edited 2012-07-06 12:55 UTC
Patents are state-granted monopolies, after all.
True. Apart from violating the concepts of free market, freedom of competition and freedom of expression, it also means patents are privilegies, thereby violating the intended egality of democracy. In a democracy there can be no privilegies, and therefore no State-granted monopolies.
More than anything the possibility of patenting genes is horrible.
As much as I dislike the current state of affairs, I can't agree on most of your points. You say balance, but the changes you propose don't seem balanced.
* Working prototype - Is a desirable feature, but not really feasible in a lot of cases. Though I would say, that a patent that is accompanied by a prototype should be granted faster than a patent without one.
* Software patents will not be granted under any possible circumstance - Even though this is tempting, I disagree with that statement in it's entirety. Limit the patent to 2 years, a requirement for the inventor to provide a working copy of the source code, source code for the autonomous module of a bigger software project and let the inventor chose copyright or patent. If the inventor chooses the patent, the source code becomes public domain. A major requirement for the software patent(an actual software patent, not a process patent as they stand today) is that it should only cover the how part. Also, a blind test should be performed on university students, to verify non-obviousness.
* No gene patents. - Not just that, I'd go even further no patents on anything that occurs in nature. Anything that is found to occur in nature is to be deemed a discovery and thus not patentable.
* patent terms should adapt to the industry they belong to - It should be very dependent on the industry and adjusted based periodically.
* Patents must be productised. - That actually clashes with your requirement of a working prototype. However, within 2 years the patent holder has to demonstrate that they either have a product or being used in products or that they have continued to actively develop the product that contains that invention. Remember, that some products have a very long development cycles. Fusion reactors will not come to pass for another 40-ish years.
* Patents should be non-transferable; in other words, they must be tied to the inventor. - Disagree, but sales of patents have to be very restricted.
* Patents will no longer be assumed to be valid by courts. - No. Patents have to be viewed as neutral documents demonstrating that the owner has in fact registered his invention and passed a certain level of scrutiny. Patent text, however, does explain why it is an invention and novel.
* Preliminary injunctions should be a lot harder to obtain. - Preliminary injections should be granted only if irreparable harm is proven. Alternatively if the invention is so essential to the infringing product, that by removing it from the product would stop the product functioning; or a number of inventions constitute more than 51% of the infringing product.
* Instead of merely weighing the costs and benefits for patent holders - One thing to add, if the patent owner has no product(or licensed product) within 6 months on the market only damages can be collected until such a product is released.
* In case of emergency, the state should be able to either temporarily or permanently nullify certain patents. - Already in place.
I will also add a major requirement - the text of the patent has to be a technical text. It should not be a legal text. Any legalese in a patent should result in an immediate dismissal of the patent. The patent text should be very clear to any person in the field and the text has to give enough information to a person of average skill to replicate the invention in it's entirety.
I'll also plug this once more.Software requires it's own IP status, not a system that was designed for writers http://jalexoid.blogspot.ie/2011/07/software-needs-new-ip-protectio...
.ie? The shame... Oh well, from Wiki page of the domain, at least they seem to limit its usage
(though seriously, some less severe restrictions of such kind maybe wouldn't be so bad; less domain hoarding)
How about: "the aggressor in court must not only prove that the patent is valid, but also that it sells a product that uses this patent."
So if company A has a patent on a working product that it sells, and gets bought by company B, then we have three scenario's:
1) B bought A's patent and production line. Patent remains valid when going to court.
2) B only bought A's patent and has its own product that uses this patent. Patent remains valid when going to court.
3) B only bought A's patent but does not produce a product. Patent is no longer productised and hence becomes invalid when going to court.
So there can be no patent trolls. Hoarding and spreading FUD/blackmailing a la Microsoft would still work but the odds of winning a court case are much lower, so the blackmailing wouldn't be as efficient.
Best way to fix the broken patent system is to have no patent system.
Copyright law is providing enough protection for software developers.
Invention are the creation of people,
so patents must be granted only to people.
In academia investigators carry their grants
wherever they go to work.
Patents should be granted same as academic grants.
The profits of the patents should be for the granter.
I suggest government should be the granter.
There are several problems with the OP's suggestions.
First, demanding that a patented item must be "productised", which I assume he means brought to market as a prototype already _is_ a "product", withing a short time frame is simply unreasonable. There are countless reasons why a product may be delayed, countless of which being out of the control of the company producing the item. You do not punish somebody who is already being victimized by circumstances out of their control.
Second, gene patents should be granted so long as the gene sequence is not naturally occurring.
Third, research & development can take tremendous investment, both time & financial. If companies are offered no protection what-so-ever until they've managed to overcome every hurdle in r&d'ing something new, they are far less likely to take the risk in the first place. You can not helping innovation, you are severely hurting it.
Fourth, there is no reason what-so-ever that patents should be non-transferable. Not every inventor or innovator is a business. Not every inventor or innovator is independent. Not every inventor or innovator wants to participate in things beyond the invention or innovating stage.
Fifth, you can't declare all patents invalid unless proven otherwise as it would impose enormous financial burden on both companies and individuals, as well as the court system which is funded by the tax payer. This would be a massive waste of time, money, and resources for everybody. The idea that you, by default, punish everybody all the way down to the tax payer is absolutely ridiculous.
Sixth, assuming judges have the knowledge & experience it takes to make sound decisions as to 'market impact' is absurd. A judge is simply not the proper tool to use to determine how a market will be affected. A single judge should not have the power to deem what's good and what isn't for business. That is not the roll of a judge. A better solution would be a panel or governing body of individuals with the proper expertise making rulings by majority vote.
To sum it up, you make some huge mistakes. You propose rules that hinder invention & innovation by placing restrictions/punishments on those actually doing or paying for the inventing/innovating to take place. You propose placing time limits that are simply unrealistic for real world application. Ignoring the fact that we live in a world based on economics & trade doesn't make it any less true.
Most will agree that the patent system is flawed and easily abused. But you don't solve that problem by going from one extreme to the other, quite possibly making things worse in the end.
Biggest problem is the drop in quality of granted patents. The system just needs a built in punishment for poor requests and the resources to sort out those poor requests. Seems that there should be a fine for patent system abuse, proportional to the abuse. That would help fund the expenses for proper screening. Also to keep the burden off of the government, we let the corporations sue each other to invalidate stupid patents, and then if found guilty the patent is revoked and the company fined ($ goes to justice system to cover court costs, the rest to the patent office.) If a company continues to get patents revoked because they're gross abusers, the fines grow quickly.
To straighten out the current system w/o all companies getting sued into oblivion, we include the option to allow companies to self-revoke their patents that are not in current contention. Probably during certain grace periods. This motivates companies to not keep a large store of probably junk patents.
Also, perhaps a bit of self regulation by bounty. Basically anyone can enter a patent and it's automatically accepted, but if someone calls you on it, you're liable for court/patent office costs if you loose and some bounty awarded to the patent bounty hunter. Otherwise the patent bounty hunter is liable for court/patent office costs for false accusations. Maybe a jury of peers would be engineers in the related field.
Or perhaps we could just have applicants pay 10k per year to maintain something in patent status for a maximum of 10 years.
A variation would have the fee double every year, so maybe 500$ the first year, 1k the second, 2k, 4k, 8k, 16k, 32k, 64k, 128k and finally 256,000$ for the 10th year.
That would at least help with the number of junk patents by either motivating companies to patent less, or by providing the patent office with more staff and resources.
Perhaps 2 classes of patents, one where a very heavy fee is paid upfront, but the patent is fully screened, peer reviewed and validated when granted. (For big ideas or things you'll license.) The other would have a lower fee, but the applicant is on the hook for penalties incurred because of an invalidated patent etc. (see my other commnet)
Disclaimer: this post is mainly based on my professional experience of a litigation lawyer in Russia. Things may be very different in U.S. and in Europe. (I hope they are.)
The part I'm less comfortable with is
It seems the judges are also becoming fed up.
http://www.h-online.com/open/news/item/US-judge-criticises-the-pate... Edited 2012-07-06 20:31 UTC
I would like to see resold patents used for defensive purposes only. Some of these Nortel patents formed the basis of a company created solely to look for companies using the patents in their products, and to sue the heck out of those companies. Might be nice to have the govvies step in; after all, they approved the sale of the patents in the first place.
About the original post: I like some parts of it, but it seems overly biased towards companies, e.g. in requiring the creation of products in X time, requiring working prototypes, etc. No patent system should close out the "small guy" or individuals who don't always have the capabiity of financial background to do so, thus would be always cornered to sell ideas to corporations.
There's no question the proposals from the OP are simply unrealistic for real world application. But, to the OP's credit, he did preface with "Let me begin by saying that I, by no means, am claiming this set of proposals is perfect, watertight, coherent, legally feasible, or workable."
Just this past week I wondered whether to reach out to you personally to understand your position on patents or bring it up in the comments to one of your articles (usually targeted at Apple) but feared the fanboy label and all the dreck that goes with defending Apple would cloud my sincere attempt to understand your position. Thank you for spelling it out, and more important, at least recognizing that SOME innovation is worthy of protection, for at least some period of time. I do fear that the world will be a worse place if a company that transforms an industry (like the iPhone for example) has no protection over the "cloners" who don't EVER innovate, but in 90 days (or less) can kick out a virtual clone of hardware, or software for that matter, but more important, does it so poorly that at the surface appearance may be similar, but once a consumer purchases it, it's only a mirage and the product is a poor clone, and more often unusable or severely crippled. Thus as you explain, patents should not only protect unique, innovative engineering, but should also consider the consumer as well. A pretty tall order, and if it was easy it robably would have already beautiful into place!
My impression is that patent trolling as just as bad outside software, so I see no reason to have it there either. Patents have consistently done the oposite of what they where supposed to do. I thinkt he fundamental idea is broken and cannot be fixed.
Intellectual property should be protected by copyright. If copyright does not cover something, then I don't think it should be protected.
Sorry for the bad English, English is not my mother tongue.
There is an easy solution to the patent problem. It is not perfect but so system is.
This solution keeps the broken patent system with have. We cannot change it easily or in the short/medium term anyway.
And that solution is: create an incentive for not attacking with patents in court.
If we add to the current laws that for each patent falsely used (irrelevant or non-valid) to attack, the attacker would have to pay $1000,000 to the defender.
This has number of advantage:
- A troll cannot use is package of patent to attack a company. It needs to be very specific. It would then be possible for the defender to check the patents and act in consequence
- Company will need to make costly research before attacking. The use of patent in court will become rare
- Since using patent to attack will be such a risky enterprise, fewer patent will be registered
I've thought about it carefully myself, and I signed up just to post this;
The most simple thing I thought of for the whole patent mess is to just price patents on an exponential basis; start patents off at a low price, $50 to actually file a patent; but every patent filed is more expensive. I've come up with an exponential curve of...
$50 + (1.03^X)
Where X is the number of patents an entity can access. When I refer to an entity, I'm referring to a company, its subsidiaries, board members or significant shareholders who would all count into the exponential price - though it doesn't necessarily mean the parent company has licensing rights to those patents.
Part of this is that upkeep would be required, just like domain names you'd have to pay yearly to keep it active (though they would still have a maximum lifespan). In addition, when the yearly payment is due, the patents a company wishes to keep have to be specifically listed. Any patents that aren't renewed become public domain.
So if Maxis had 50 patents and EA (owner of Maxis) had 200, Maxis would pay dues for 50 patents ($2612.80/yr) and EA would pay dues for 250 patents (66440.71/yr)
I would still allow entities to transfer their patents, however the lifespan of the patent would be halved every transfer. This would include the situation of two companies merging or one company being bought out. The only exception would be if a patent had less than five years until expiry. E.g. A subsidiary could sell its patents to a parent, but a patent with 20 years would be halved to 10. A patent with 8 years would be almost halved to 5. This would still allow a great idea to be sold, but it would prevent ideas from being hoarded because those ideas effectively halve in value every time they're sold.
Since the fees are much lower a small company who is struggling to produce their patent can afford to hold onto it if it's really worth it while building a product. On the other hand, if you're a patent troll it would become difficult to own enough patents because of the cost of owning large numbers of patents. Patents would eventually hit a threshold and become too expensive for a company surviving on lawsuits.
As an example;
A small innovative company is not likely to have more than a few patents, or even a couple dozen if they're busy beavers. If they had 25 patents, their most expensive patent would cost $54, and they'd be paying 2721.70/yr to keep those patent from public domain.
If a large company like Google had 400 patents, their most expensive patent would be $132,500/yr and their yearly payments would be $4,567,423/yr.
If a company is registering a 469th patent, that is a million dollar patent. If you are Microsoft or HTC or Google and you really believe your patent is that valuable you will make the hard decision to either pay that much or dump less valuable patents.
Overall, it keeps it open to individuals who could easily afford those small fees, and ensures companies responsibly choose which patents are actually valuable. If a company with clout feels it can make a billion dollars off a great idea, spending $1,000,000 on a 469th patent for even a year makes sense. Also, the threshold for having enough patents to establish an effective patent-trolling firm is too steep; the 500th patent would cost 2,500,000/yr.
In my opinion, I look at each patent like it should be the next million dollar idea. That's what patents are for, protecting innovation to help the common man unleash those brilliant ideas, and help big companies protect themselves from millions in R&D being stolen.
Microsoft hit 1.2 billion dollars in sales and it owned only 5 patents. I think my formula is actually too forgiving, I'd adjust it so 50 patents would begin to be pricey. If something like this was implemented now, a very conservative formula could be used and every year be tightened until it hits its final number. Companies would be forced to gradually sift through their ridiculous portfolios and weed out the worthless materials until eventually settling on the real inventions and designs they should actually be proud of. But even my forgiving formula will never see the light of day, things will just stay broken - and that's just sad.