Legislation To Overhaul US Patent System

FutureDomain sends us to a PC World report on the filing yesterday of legislation to overhaul the US patent system completely. The US has the only system worldwide that tries to ascertain who first invented a thing — everywhere else the criterion is who filed first, and the new legislation would bring the US into line. Identical bills were introduced in both the House and the Senate by, in each case, bipartisan sponsors. The corporate roster of backers includes Microsoft, IBM, Amazon, and nearly everyone else. From the article: "The provisions of the Patent Reform Act would... restrict damages that patent holders can receive for infringement lawsuits, create a new procedure to challenge the validity of a patent after it has been granted, and boost resources for the US Patent and Trademark Office."

Re:First Post!

[Anonymous Coward]

I mean "Prior art!"

Nothing to do with prior art - prior art invalidates patents in first-to-file countries* as well as first-to-patent countries.

It's more when two inventors develops something novel at the same time. Who gets the patent? First to invent - or first to file?

* Which include the Phillipines as well as the US?

Re:Translation ... Garage inventer will be screwed

[Carewolf]

No. Prior art stays the same. If anyone has published a the same idea before it is not patentable (in theory, in practice USPO will still not check it),

The old system just had the option of companies not filing patents and not publishing their results could come in and claim they invented the idea first.

The new system is much less corrupt and more open.

Re:Frankly, this worries me

[Anonymous Coward]

Under current US patent law, keeping detialed records of your designs and inventions is only useful if you intend to patent them, and then need to prove the invention date to get priority for your patent application over some other guy's application. Detailed records don't do squat if you don't file patent applications for your work.

If you aren't filing patents, then the only way to protect yourself in the current system is to publish your work (thus making it "prior art"). Regardless, someone could later file for a patent and claim that they invented it before your publication date.

In the "first to file" system, publication is still the way to keep yourself safe. Published inventions cannot be patented. What's better in the "first to file" system than the current "first to invent" system is that once you have published something, then it cannot be patented any more, unless someone managed to file their patent application while your publication was still in print (thus getting a filing date before your publication date).

The key point is, you need to publish your work to be safe from patent trolls.

Re:Wouldn't this actually be a huge step BACKWARD?

[JTL21]

In the rest of the world published prior art still disallows patents.

Filed first is just the decider when two applications are made for the same invention (which was secret until the applications).

In most of the world the person who gets to the patent office first gets the patent and the one who gets there second gets nothing. In the US you can claim that you invented it before the other party and try to prove it with dated notes etc. which were previously not published. In theory the first to have the invention gets the patent not the first to file. This leads to messy legal fights.

In the rest of the world the incentive is to file (and therefore publish) first which gives a clearer decision on this issue.

Joseph

Re:I'm not sure I get it

[Bastard of Subhumani]

The advantage of the 'first to file' rule is it prevents the situation where A patents something and B goes "Oi! I thought of that last year. Honest. Ask my cousin C and pal D". Big companies have an advantage in that kind of trick: 'Sure we did, just ask employees E1 ... En.

It has no bearing on patenting beer, the sky or sex, which ought to be covered by other rules (novelty and non-obviousness to start, but also the fact that they naturally exist). Or at least it shouldn't have... this is the USPTO we're talking about.

Re:Pork for the big companies

[thebdj]

Each of these companies is well-known for abusing Software patents; the first two in order to maintain their monopolies.

Can you please tell me what monopoly IBM has? Are they a HUGE multi-national corporation? Yes, but a monopoly they hardly are. What share of the PC market do they own? Oh right, all that was sold off to Lenovo. What percent of the server market? Unix Market? Now, what percent of the OS market is Microsoft? Go re-read the definition of monopoly first. As for abusing patents, Microsoft and IBM have not sued anyone with their patents that I am aware of. In large part because the computer industry would be destroyed. There is sort of a *wink*wink* agreement within the computer industry as it relates to patents. I will give you that Amazon is basically abusing their really shitty patent for 1-click.

Usually "reform" by these folks means "give me a bigger piece of the pie in order to stifle innovation and keep the little guys out".

Actually, some of the reform these guys have proposed and the briefs they have written for SCOTUS cases have been in favor of things that would REDUCE the number of patents. At least one item would probably destroy half of their patent portfolios. It is actually kind of funny, but the people who always seem to be on the opposite side of the technology companies in these sorts of reforms are the big pharmaceutical companies. Makes you wonder who is the bigger evil.

Pardon my suspicions, but I doubt this so-called "reform" is here to help us. TFA isn't clear on this, but it sounds like this is just to establish the first-to-file basis of patents, rather thna first-to-invent. First-to-invent is advantagous for the little guy, so it's no wonder the big companies are behind this dubious "reform". This is not good legislation.

First to invent is a mess. Speaking as a former patent examiner, it is not pretty and creates about as many issues as it would ever solve. It is out-of-line with the entire world, which allows for other countries to issue the exact same patent to another company than who it was issued to in the US. It also allows companies to come back to back-date their date-of-invention to circumvent prior art, something they would no longer be able to do with the current system. It would also get rid of interferences, for the most part, something that takes up the BPAI time instead of working on appeals, which is far more important to the overall process.

If they really wanted to reform the system, it would be better to either toss out Software Patents (and harmonize our system with much of the rest of the world) or at least require a working prototype before a Software Patent is granted.

Um, working prototype is currently a requirement FOR ALL PATENTS. You do not patent an idea, you patent an invention, the actual item. It is actually within the right of the USPTO to ask for a working prototype if the functionality or existence of an item comes into question. This is very rare, but it can be done. So, for a Software patent to be issued, it would actually have to be in some form of use. Removal of Software patents and more importantly business-method patents is important. It would truly create harmonization with the rest of the world, since the US has failed multiple times to convince other nations to allow patentability of software.

Re:Translation ... Garage inventer will be screwed

[MrNJ]

No. Under the current system, the first to invent has 1 year (I think) to use the invention without worrying about "race to the Patent Office" that exists in the "first-to-file" countries. If anything, changing the system to the "first-to-file" rule would result in more filings because companies would be pressed to file as soon as possible to establish priority. Instead, the Congress should clarify the non-obviousness requirement and perhaps require USPTO to hire examiners with experience in the software development to review software patent applications.

Not a prior art problem

[PatentMagus]

The same prior art applies in a first to file system. If it is out there and published/used/known, then it is prior art and can be used to reject patent claims.

Trade secrets, however, get a good hosing. If you keep something secret and someone else patents your secret, you could suddenly find yourself paying royalties.

Re:Pork for the big companies

[kidtruth]

"Um, working prototype is currently a requirement FOR ALL PATENTS. You do not patent an idea, you patent an invention, the actual item." Incorrect. You must only reduce your idea to practice, which means you can make a detailed list or explanation of your invention rather than a working prototype. You can patent something that doesn't exist, or doesn't work. It only must THEORETICALLY work. Of course, if the language is not extremely specific as to what the invention does, the patent claims will not protect against anything. This is a protection of the USPTO to protect those who don't have millions to spend on R&D.

No more prior art? I think not.

[vtcodger]

***No, no more prior art. If you filed first, screw prior art. It's yours.***

I'm pretty sure that's incorrect. Prior art should still hold in that you can't patent something that was described by DaVinci, Ben Franklin, Alexander Graham Bell, Erosthanes, or some dude in Ohio in 1998 on his web site. What first to file means is that the USPTO no longer needs to flip a coin when it gets three applications for the same thing and needs to determine who made the invention when.

See http://www.heise.de/english/newsticker/news/86141 [heise.de] which makes if pretty clear that Germany -- a first to file country -- considers prior art in judging the validity of patents.

Of course, I'm no patent expert, and the US Congress with a little help from the lobbiests is capable of coming up with absolutely abominable legislation . But I'd keep an open mind on this one at least for a while for a while.

Re:Bizarroworld

[Anonymous Coward]

The problem is that patent law is very complicated, and you need significant experience to write and litigate patents. Writing patents can be done on a budget, but don't expect to get an enforceable patent through the PTO on less than $8k. Also, the cost of patent litigation is quite high, however, if you have a good patent and targets with lots of money, then you can likely get a patent lawyer to take it on a contingency basis. However, the regular legal fees for a patent case are almost always in the low seven figures if not done on a contingent basis.

Re:First to file

[ip_vjl]

First to file is the method in use in countries outside of the US. It doesn't mean that prior art is irrelevant. It simply means that if a patent application comes in that is valid (new, non-obvious) it is granted to the party that submitted the application first. This is unlike the US where I could file my application a month AFTER you, but if my lab notebooks show I was working on the invention before you were working on it (even though we knew nothing of each other) I would be awarded the patent.

Prior art still comes into play in the initial evaluation of the patent. It can also be used to challenge a previously awarded patent. Going to first to file would just put the US in line with the method used by the USPTO (and just about everywhere - if not everywhere - else in the world.)

The only problem you would face is the typical Slashdot view that disclosures don't matter, as the patent examiners will "rubber stamp" anything they get - but part of that depends on how you publish your disclosure. If you put it on your own website, it is unlikely to be seen. If you publish with one of the high visibility journals - you're more likely to be seen.

(Full disclosure: I used to work for IP.com which does this sort of thing. So I've had a fair amount of exposure to the workings of patents and disclosures.)

Re:Pork for the big companies

[NearlyHeadless]

http://www.uspto.gov/web/offices/pac/doc/general/i ndex.html#model [uspto.gov]

Models or exhibits are not required in most patent applications since the description of the invention in the specification and the drawings must be sufficiently full, clear, and complete and capable of being understood to disclose the invention without the aid of a model.

A working model, or other physical exhibit, may be required by the Office if deemed necessary. This is not done very often. A working model may be requested in the case of applications for patent for alleged perpetual motion devices.

When the invention relates to a composition of matter, the applicant may be required to furnish specimens of the composition, or of its ingredients or intermediates, for inspection or experiment. If the invention is a microbiological invention, a deposit of the micro-organism involved is required.

Re:Legalized theft!

[zymurgyboy]

No. OS/PD projects are copyrighted not patented. Whole different animal.

Re:Bizarroworld

[N8F8]

The actual costs for properly documenting a patent are MUCH higher. IANAL and YANAL so look into it sometime. And even if you get one you have to be able to spend the money and time to defend it.

Re:Legalized theft!

[yfarren]

I have mod points, and was looking for some-one to point out that the parent and grandparent are not insightful, just wrong. There really needs to be a -1 wrong Mod.

Anyhow. This legislation makes it much MUCH easier for open source projects. Instead of having to write something, and then find out someone else claims to have invented it, and they filed a patent before you published, you just have to search filed patents. If it isn't there, publish your general idea, and write your open source project.

Similarly, both the parent and grandparent seem to totally miss that you most definitely cannot patent something from an open source project, as, anything which is available as open source, is already published.

Here is the thing. Any public disclosure (prior to filing) invalidates a patent. Now, I am allowed to talk with a friend, to further developement, but, to cover your ass, you really should get him/her to sign an NDA before talking about it, because simply talking about it constitutes a public disclosure. Proving that public disclosure can be an issue.

Now, what this legislation does is, it says that between 2 people (corporations, groups, committees, whatever) who have the same idea, and both try to patent it, provided that neither has published anything, and it is a useful, non-obvious idea, we will give the patent to the one who files first. This eliminates confusion about who should gt it, with both parties "no, I had the idea first. So, for the most part, to avoid infringement, you just have to search relevant patents. There is still some time, between when a provisional patent is filed, and when the USPTO accepts (and publishes) it that you could have various "submarine" patents get into an open standard, or your open source project, but this legislation doesnt make it possible, at all, for some-one to take your OS project, and start filing patents based on it. Parent and GP are crying wolf, without understanding what they are talking about.

Re:Legalized theft!

[justzisguy]

Alright, more digging. From the Patent Reform Act of 2007 [house.gov],

102. Conditions for patentability; novelty
(a) Novelty; Prior Art.—A patent for a claimed invention may not be obtained if—
(1) the claimed invention was patented, described in a printed publication, or in public use or on sale—
(A) more than one year before the effective filing date of the claimed invention; or
(B) one year or less before the effective filing date of the claimed invention, other than through disclosures made by the inventor or a joint inventor or by others who obtained the subject matter disclosed directly or indirectly from the inventor or a joint inventor;

With the exception of the inventor publishing something and then filing for his patent in less than one year from publishing (e.g., publishes a paper with the IEEE), the public domain material would qualify as prior art under (a) (1).