Slashdot Mirror
Scientific American On Bad Patents
dltallan writes: "Scientific American has a short article in which Gregory Aharonian presents his picks for the four worst patents granted. I like the patent for training with manuals (1998)." The Bustpatents site is worth spending some eye-rubbing time on.
You know, I wonder if this wouldn't be a good way to fight the RIAA. I remember a quote from the guy who leads the patent office saying he thought legal arguments should be patentable. His argument was basically "Why the hell not?" as I recall.
Now, I don't think that you can patent legal arguments at this point in time, but you can patent business models. I mean, we're all smart people, I say why not just patent everything we can think of that the riaa might need technological to fight fair use, etc. I mean they already have things like digital watermarking, and some business models, but why not try to stay ahead of the curve? I remember thinking about a Digital Rights management system built into an OS a couple of months before Microsoft's patent was announced. If I'd thought of it years ago and patented it I would have stopped MS from ever implementing it in the core of the OS. Hrm... not that I wouldn't have been unbribeable, but, maybe this could be done through some non-profit organization like the EFF or something.
Hrm, maybe I should patent this idea... method stopping tyranny through abuse of the patent system...
autopr0n is like, down and stuff.
Part of the problem with the patent office is that technically minded people don't want to work there. Sure you get to learn about lots of creative ideas and there is good deal of job security, but who really wants to work for the patent office? As far as I know there has only ever been one really famous patent clerk, and he left shortly after that paper was published.
People with technical skills and training to solve problems aren't interested in a job that is largely research and paperwork. Who can blame them? At the University of Maryland (just outside DC), some people from the patent office came to give a well publicized lecture about their trade and next to no one showed up to hear it. The most knowledgable people will go off and invent or even just teach and then people who can't find positions elsewhere end up working for the USPTO. [I'm sure there are some bright patent examiners, but they sure aren't likely to attract many.]
One might try using pay and perks to make the job more attractive, but that's only likely to go so far. It occurs to me that we might do much better if there was a system resembling that of peer review for scientific research. A lot of problems might be solved if patent applications were reviewed by a pool of people that had recieved patents in related fields. Just add a priviso that people that recieve a patent are obligated to review 3-5 patent applications per year for the length of the patent. Remove all the identifying details and send each application out to several people to referee.
Of course there are lots of details to work out, but IMHO such a system could go a long way towards improving the current state of affairs with the granting of patents.
But they would never have to go as far as to the court (BTW, I've never understood the American court frenzy either). Let's say that company Foo.com submits a patent application for no-click shopping. Upon submission, the PTO verifies that the claim is potentially patentable (i.e., it's a formally valid and sensible claim). The application is acknowledged and made public. Then another company Bar.com sees the application and submits proof that no-click shopping has been used well before Foo.com. The PTO (not the court) verifies the claim and invalidates the patent.
The PTO would not have to find prior art itself if it could tap the resources of the community for that, otherwise it would make decisions on the same grounds as today. Also, considering the extended processing times of recent patents, especially IT patents are either obsolete or then overtly general in scope once they come out of the mill (independently of whether software should be patentable at all). The kind of shortcut I was suggesting would cut the processing time to minimum, if the patent could be immediately exercised until contested.
Ok, I'm a scientist, and on that field the process has proved to be very effective. Someone claims something, and then everybody is free to try to prove otherwise. Every hypothesis is held potentially true until falsified, but nothing is ever proved true, only the most plausible/widely accepted theory.
Comic-not
Existence usually comes as a surprise (Idem)
I think the USPTO should allow a time for public comments to be issued on pending patents before the research process begins. Perhaps via some easy to use Internet/Web interface. This would certainly reduce the work load of these guys, and pre-sort the bad/invalid patents out. Perhaps the ones with lots of comments would receive more attention by the USPTO as opposed to pending patents that recieve little or no comment. By essentially envolving industry in the process it is actually made more efficent. I suppose a possible draw back would be someone trying to slow down the approval of a good patent by making false claims, but a method could be devized to reduce or eliminate such behavior. Maybe the USPTO should install Slashcode and allow pending patents to be modded up or down depending on their worthiness. Regardless of how it is carried out, something needs to be done to change how the system works because it clearly doesn't work now. One idea I had would be to follow up on companies to see if they actually build/sell the patented product within a given time frame. If they don't then the patent would be lifted allowing anyone to use the information. Wouldn't that help a lot more mousetraps hit the street rather than being tied up forever in some government stack of paper work?
This is something that never seems to get much attention. People think as long as someone makes money from it, it's a good thing as long as they don't make too much money...it almost seems like as long as someone can make money from a thing that thing provides employment...but that's not always the case.
:-) But at least I'm not going to claim that somehow we don't need people to provide the resources for everyone else, or that somehow we can have the state nobly and objectively manage them all...
It's important to distinguish the exchange of labour for money from the rental of a good for money. Rental always ends up being money for nothing at the end of the period. Think of all the means of maintaining a large fortune(and therby a power structure) and they all involve rental in some form. This is why it requires money to make money. The traders who make money out of markets are those who have the fortunes to swing them and the cash reserves to withstand losses. Employment itself is essentially a system where those who aren't rich rent a factory in order to convert their labour into cash.
If it was impossible/illegal to rent it would be impossible to become or stay rich.
Ofcourse this is all Red Commie crap
Marx was always better at critiquing than providing any solution.
The US patent office can only search things in the library of congress or their library (i.e. stuff that is patent pending or has been patented (they discard the stuff they reject)). They can't do a google search for prior art since that would leak info to google.
Because most of the inventors of the cool computer stuff didn't think their work was patentable, they didn't submit it and now if you can find something that hasn't been sbumitted, you too can get a patent on something you didn't invent.
The same is true for business practices which can now be patented. With no "prior art", anything can get a patent.
One easy fix to this would be to do something as patent a method to something uesless (but patentable) and submit all of google's database. the problem with this approach is that it requires the entire database.
Wow. That got unfunny about 2/3 of the way through the first sentence. Then it dragged on for several more. By the time I was done reading it I was so bored I wanted to gouge my eyeballs out just so I wouldn't have to read another one of your posts again.
Is there any mechanism by which members of the public can make available to the PTO examples of commonly used, prior-art software, prior-art biotech, prior-art whatever takes your fancy? If not, I think that is something that is necessary - we can moan all day about the PTO granting patents on "obvious" stuff, but perhaps we need to give the PTO a list, to help make their job easier. As another poster pointed out, a PTO clark can't do a Google search for the details of a new patent application, as that leaks information to Google.
What I'm thinking of is a place where I could describe, say, shopping on a website (to pick an example from bustpatents), and describe in very general terms how shopping on a website works (e.g. "You fill out forms in a web browser, and submit what you want to buy along with your credit card number preferrably over some form of cryptographically secured link such as HTTPS"). It needn't be an especially good example, and (I would say) the more general the better.
The PTO may well require the information to be public, which makes sense (no-one would post anything of value on it). And perhaps the PTO could somehow penalise applicants who apply for something already described on this list (or on a "small" number of other lists)
I can see that it would need a form of moderation system in order to avoid abuse, as I'm sure the PTO probably don't want a few thousand "common examples" of spam. But unless we the software community describe (preferrably in a concise manner, and definitely in one place that the PTO can grab a complete copy for its own searching) what stuff is obvious, the patent office doesn't stand a chance.
PS: I live in the UK, where software patents are not legal. However, you'd be amazed at the number of "mechanisms" that most people would describe as "software"
This is from an ex-GS-1224 (Patent Examiner)
The issue of quantity vs. quality has been a long standing issue in the PTO; it was a raging issue back when I joined in 1973. In the "olden days" (say up until the 1960's, according to the then lore) an examiner could make multiple rejections against an insistant applicant; the ethic was to reject claims. Also, the examiner's judgement was given high authority, and applicants had to make a good case to overturn adverse actions.
Well, the patent bar got pissed and started to get more aggressive about challenging examiners; after all, there was abuse of the system, and it is not much better to reject somtthing over clearly irrelevant prior art then to pass out a patent over a clearly anticipating reference. Also, applicants were concerned that applications were taking several years to get issued, making the whole process too expensive and time consuming.
As a result of these pressures a new bred of management was cultivated to counter this. One of the key provisions of this effort was the creation of the "Compact Prosecution" policy. This said that a typical contested application would receive a first action, then, on response by applicant, a second action, either allowance or a "final" rejection. If finally rejected, the applicant could appeal, but was not entitled to have any pre appeal submissions to be considered except under limited circumstances.
To "encourage" examiners to to follow compact prosecution the perfomance evaluation system for them was changed to significantly count an examiner's "production" by giving a credit for the first action and for the disposal of the application (generally abandonment or allowance).
An examiner still has discretion to make actions after the first non-final, but, clearly, the incentives go greatly against this.
From this beginning, management found an easy metric to rate examiners:very "objective" and requiring little "evaluation" of the quality of the work. Patent examining involved patent application "processing" Real quality only occured by hapenstance such as protest from the outside, but low production was the basis of a ticket out.
As far as the issue of software and business patents, look to the courts for that one; the Office, in true bureaucratic inertial fashion resisted those applications, but the patent bar won in the old Court of Customs and Patent Appeals and its post 1982 sucessor the Court of Appeals for the Federal Circuit. An express change in the Patent Law will be required to change this (or some kind of radical change in the legal thinking of the judges on the Court).
Several US companies own patents for individual human and animal genes. No, not modified genes. Naturally occuring ones, like the one which causes Cystic Fibrosis.
If you're developing a CF test which looks for this gene (even from scratch), you will be paying patent royalties.
It's logically consistent with this that you could be paying royalties if you (or your child) has CF...
The only thing you can accurately describe as "Scotch" is a sticky tape made by 3M. And it's
US Patent 6025810:
A method to transmit and receive electromagnetic waves which comprises generating opposing magnetic fields having a plane of maximum force running perpendicular to a longitudinal axis of the magnetic field; generating a heat source along an axis parallel to the longitudinal axis of the magnetic field; generating an accelerator parallel to and in close proximity to the heat source, thereby creating an input and output port; and generating a communications signal into the input and output port, thereby sending the signal at a speed faster than light.
Certainly not going to get shot down by prior art...