IEEE Proposes New Class of Patents

cheesedog writes "The IEEE Spectrum proposes a new type of patent that wouldn't require formal examination, would cost significantly less than traditional patents, would last only 4 years from date of first commercial product, and which wouldn't carry a presumption of validity. These 'limited patents' would be attractive to innovators in the fast-moving high-tech industry that can't wait 18-24 months for patent approval, and would help improve patent quality by populating the USPTO's prior-art database more efficiently. Additional commentary on this proposal is available."

Great idea!

[timjdot]

This is a very good idea. My current patent application was filed Jan '01 and is still being reviewed!

TimJowers

Re:So

[Eightyford]

...but is giving out patents without formal examination really the answer?

I think so. This just establishes prior art in a very clean way. If the patent shouldn't have been granted, the courts can come into play. Anything is better than the system in place now.

Re:Improve patent quality?

[timjdot]

Check out http://www.shouldexist.org/ [shouldexist.org]. ShouldExist is a superb place to anti-patent ideas. I will not be surprised if a fair number of software patents being filed have been already mentioned on ShouldExist.

TimJowers

Re:without the presumption of validity?

[morgan_greywolf]

Exactly.

Imagine Acme Corp. files a 'limited patent' for Widget X and it's granted without examination. My widget, Widget Y does the same thing as Widget X, and it's actually got a real patent pending and has been on the market for 3 years.

However, Widget Y hasn't been selling very well due to my inability to market the product, and well, I can't afford good legal representation. So I can't sue Acme Corp. at all... worse, Acme Corp. notices my product and decides to sue me! Since Acme got their 'limited patent' first and mine is just pending, Acme wins!

Screw that. It sounds like a patent abusers' wet dream.

Re:So

[alicenextdoor]

This has already been tried, in Australia. In fact, the law won an IgNobel prize for John Keogh and the Australian Patent Office [improbable.com] for patenting the wheel in the year 2001. Apparently he did it to demonstrate that the new patent laws were pointless. I have no idea if his patent has ever been challenged in court!

Just limit the number

[GrEp]

Mandatory licencing via a maximum 15% tariff on patented goods would solve the problem.

Right now the little guys get eaten alive having the burden of getting their goods to market without the propper resources to do so. With the 15% tariff all they need to do is patent their invention and the market will reap the profits for them.

Yes, big pharma might get lowballed for their R/D costs, but on the bright side they wouldn't have to spend billons on those drug ads that appear during prime time.

Also, software patents wouldn't be a problem because paying an extra 15 cents on your $1 software download isn't going to hurt much.

Re:So

['nother poster]

But some poor bastard, erm I mean citizen, in Australia shouldn't have to pay money to barristers to plead a case to overturn a patent, even a minipatent or whatever they call them, so that he can make an item based on a 5000+ year old concept. I know the wheel patent was done as a protest, but plenty of other stupid ideas are done with the express purpose of extorting money from innocent people in a legal manner. Patents are important, but the law in Australia, and the proposed law in the U>S> are asinine and harmful.

Re:So

[x8]

What if the patent office had a link on each patent's web page that would allow members of the comunity to submit prior art, triggering a formal review? (Is this possible now?)

I can think of a few benefits of this:
1. The patent office could focus their formal review efforts, gaining efficiency.
2. Avoids hiring a lawyer every time you find a patent that obviously should be invalid.
3. In the proposed IEEE system, having a lot of prior art patents would make it easier to point out prior art.

Complain, Complain, Complain

[DavidD_CA]

Everyone on /. consistently bashes the current IP/patent system and the PTO, always offering heaps of reasons why it sucks.

This guy, an actual *professional* in that field, comes up with some ideas and spends the time to think them through, document them, review them with peers, and even has a huge body (IEEE) behind him.

Yet 90% of the posts on here are negative, with absolutely no substantial suggestions that would improve the situation.

RTFA

[Dr. Donuts]

I see a lot of comments from people that obviously didn't read the article. Some have made some good points, for those that did.

After reading the article, the solution proposed is a good solution IMO. A patent granted through this system solves many of the problems with the current patent process.

1. Eliminates the upfront burden on the patent office.
2. A patent is only enforceable if a commercial product is produced by the patentee. That's a biggie. This gets rid of the current batch of "IP" companies, whose sole existance is pumping out patents and litigating. No product, no enforceable patent.
3. The first step in any litigation is an examination by the USPTO.
4. These mini-patents can be challenged via prior art with a small fee.
5. It will help build up the prior art database, which should allow the USPTO to be quicker in making prior art determinations in any examinations they have to do.

Although people might bemoan the fact that this would grant a lot of trivial patents, those same trivial patents also mean that prior art has been established. Those trivial patents are meaningless unless a commercial product is produced. And if it's trivial, then chances are it will be easily refuted upon a full USPTO examination. Which means you've now established both prior art and refuted the patent, both of which can be referenced by the USPTO directly for later patent examinations, which should help them make determinations for prior art quicker.

Since the fee involved with one of these patents is small, it will also mean that people would be free to submit patents for the sole purpose of establishing prior art, even if they have no intent of bringing a product to market. This could be used to actually prevent folks from taking out trivial patents in the future, making a product, and then trying to enforce it, since a prior patent had already been issued. Even if that prior patent is not enforceable, it still establishes the prior art, thus invalidating the subsequent patent.

Would love to see someone's reasoned arguments as to what the pitfalls might be under such a system.

Re:Improve patent quality?

[Anonymous Coward]

Not really. shouldexist.org seems to publish mostly unimplemented ideas. But you can't patent ideas, you have to implement it before you can get a patent. Therefore, all the unimplemented ideas on shouldexist.org are still patentable by someone who goes through the effort of working out the details by creating an implementation. As other have said, the proposal requires the patent owner to distribute a product, which itself requires an (economicly feasible) implementation.

On the other hand, it is true that to establish prior art you need to publicize your implementation, and there may be good ways to do that other than paying the patent office. A sticking point is establishing the date. For example, you could release an open source implementation, but I'm not sure how you would establish the date of release in a way that could not be forged in the eyes of the patent office (just back-dating copyright 1996 on the source code won't do it). Maybe if you get detailed reviews in a reputable print magazine, or distributed on a dated periodical CD by a reputable third party. The more standard academic way is to publish an article in a reputable peer-reviewed journal, but that may be difficult for more trivial inventions (one-click purchases, etc.).