House approves major overhaul of patent system
news.yahoo.com
news.yahoo.com
This is a huge red flag.
"Opposition came mainly from groups representing independent inventors, small businesses and academics."
Which is not to say that you might not be right to be worried.
Or it could water down the standards even further as undertrained people come in and rubber stamp whatever's in front of them.
In general "first to file" makes sense to me because it removes uncertainty (who filed first is indisputable, determining who invented first requires an expensive lawsuit).
But what if I invented something, published it, don't intend to ever file a patent for it and someone comes in and files a patent for that invention?
If the can get a valid patent on my invention, it seems wrong and open to abuse (what if they didn't actually invent it but just filed for a patent based on what I published).
If they can't get a patent, then "first to file" seems meaningless.
Your published prior art blocks them.
If they can't get a patent, then "first to file" seems meaningless.
First to file is really about the windows between filiing and inventing. The argument for first to file is generally what you state -- that it helps clear up ambiguity.
The argument against is usually small businesses who believe that they may not patent (because they're expensive) until they have something in practice. So imagine you invent some new high efficiency light, and spend the next year getting it working. And once it works and you're convinced of it you file. But you find out that GE patented it a week earlier. But only discovered it a week before that and hasn't made much progress on it (it's only been a few weeks). You still lose the patent, despite the fact that you may be able to prove that (a) you invented first, and (b) have spent the past year implementing it.
With that said, the first to file vs invest, IMO is a small part of the reform. The bigger part is the grace period, which now has the following change. In the past you could build an iPhone, ship it, write a paper about it, and file a patent within the next year. Now you can write a paper about it, but can't sell it. Once you sell it, your grace period is over.
At least that's my IANAL understanding. And the other big thing in it is they've expanded the timeline of challenging patents. I'm surprised the EFF doesn't have a more coordinated effort to challenge patents when granted.
How does it block them? Does it or does it not matter who invented the thing? I am incredibly confused about this legislation.
As defined in the bill, prior art can be "patented, described in a printed publication, or in public use, on sale, or otherwise available to the public before the effective filing date of the claimed invention".
It's not a great time to be making laws that depend on a formal distinction between published and unpublished, since that distinction has been just about completely smoothed over by now.
What it really does is remove the option of protecting something as a secret. You have to either publish or patent (which is basically a state-sanctioned form of publishing with monopoly benefits). It completely changes the calculus of small inventors, who will now be forced to file multiple provisional patents ($100+ a shot plus time spent) and then within a year decide to shell out $5K++ to have any chance at all of not being beaten over the head by corporate patent trolls.
Not good.
That's been true for decades under existing law.
> The earlier user is immune from infringement of the patent under the "prior user defense", though.
This will be a new feature of the law under the new legislation.
For example:
Lab A and B are all working independently on a project.
1. Lab A files first and receives a patent. If without reading the patent or knowing of the details of the invention, Lab B manages to re-invent the same technology they will receive a patent as well.
2. Lab A invents first, but doesn't file. Later Lab B files for a patent it wins one as Lab A's work hadn't been described via the patent office. Lab A can still file for a patent if they can prove that their invention came first.
Essentially it seeks to turn the patent from a monopoly grant to a cartel grant, thereby rewarding invention regardless of when it occurs.
If a patent holder wants to ensure protection for their work, they need to file first and extensively publicize their filing as such no one can claim spontaneous re-invention as is the case with many software products.
In other words, billion dollar companies get all the SW patents.
HP can "extensively publicize" for almost nothing by piggybacking on other things that they do. Small companies and individuals can't.
IBM, for example, has numerous publications that would qualify. (They send them to every CS department library as well as other subscribers.)
What? You don't keep up with the literature?
I used to look at every tech report that Stanford CS received. That became impossible in the late 80s and things have only gotten worse.
All? No, they only get patents for those things that they invent.
It cuts both ways, the bigger you are the more liable the newly empowered patent bureau is to say that you could have easily found the existing patented invention.
On the other hand, since multiple parties can hold patents and subsequently re-licence them it will be easy for a a company, large or small, to protect other parties from patent trolls.
The something obnoxious like the famed "one click" patent would have multiple holders, and any could simply give a universal grant to it.
In the worst case scenario, which I don't think is plausible then the current status quo is maintained for some patents. A large company invents, files it, then publicizes it.
P.S.
I think that business form patents patents like the one-click should be null-in-void in any country.
> All? No, they only get patents for those things that they invent.
As I pointed out, the "extensive publication" requirement is a significant obstacle for everyone else, so they won't get SW patents. Thus, all of the SW patents go to big companies.
> The something obnoxious like the famed "one click" patent
"obnoxious"? That's your objection to one click?
What's wrong with novelty/prior art and non-obvious?
In every other field, recognizing that there is a problem and patenting a broad-swath of mechanisms to solve it is considered good. Why is software different?
And yes, I'm familiar with James Bessen's work. I introduced his EE380 talk at Stanford.
As long as they exist, they'll need to be "fixed".
EDIT: Computer folks tend to see the world through a particular set of glasses. They see a world of inventors in their basement you can invent with little capital investment. That's fine and is great for innovation, but the whole world doesn't look like that. If you're in an industry where the process of invention is extremely costly, patents can often be the only way for smaller entities to participate. Say you have an idea for a new radio technology. It'll cost you millions of dollars to build a prototype (the associated parts and equipment are very expensive). And once you have it --- what do you do with it? Go into production and compete with AT&T and Verizon? Patents allow meaningful seperation of invention from productization, and that's something that hugely benefits smaller entities.
"[T]he bill would put the United States under the same first-inventor-to-file system for patent applications used by Europe and Japan. Currently the country operates on a first-to-invent system" ...
"John Conyers [...] said it would permit the Patent and Trademark Office to award a patent to the first person who can win a race to the patent office regardless of who is the actual inventor."
e.g. you really don't care about patents. Guy later patents what you're doing and sues you. You prove prior work, suit fails.
Best of both worlds?
Undisclosed prior art is not a defense in first to fiel systems
The difference between one of these and a patent is best summed up by one of my favorite Terry Pratchett quotes: "It's the difference between using a feather and using a chicken."
However, the House version of the reform bill expands the defense to cover all types of claims; thus if it survives reconciliation with the Senate bill it will likely become a viable defense.
Wow, does this mean everyone who wants to not get patent trolled has to file immediately?
It sounds like looking for new gizmos without associated patents and then patenting them as yours would be an ideal strategy.
The alternative? Patent every gizmo yourself before release. Neat. A tax on innovation.
All this changes is how it is handled when two or more inventors have overlapping inventions.
But, I am not a lawyer.
Hopefully this is a little loophole that can be widened to allow broader attacks on business method patents.
"There was also strong opposition to a provision that allows financial institutions to challenge patents issued on business methods, such as ways to process checks. The opponents said the provision amounted to a bailout for banks, but Rep. Robert Goodlatte, R-Va., chairman of the Judiciary intellectual property subcommittee said business method patents, a fairly recent phenomenon, were "a fundamental flaw in the system that is costing consumers millions each year."
"An amendment to remove the section concerning the business method patents was defeated 262-158."
So I guess it's in there? Confusing.
But what about software patents? Can we discuss reform specifically related to those?
Ideas:
1.) Shorter periods of ownership/proprietary rights. 2.) More detailed applications. 3.) OR abolishment.
=============================
On the other hand, image this:
- I created an invention.
- I worked my ass off saving up $5-10 thousand dollars for patent and legal fees.
- I spend hundreds of hours of my life searching for prior art on the internet and in publishing magazines. I think I'm safe.
- I file the patent.
- Some random ass guy in Minnesota comes up with prior art and I don't get the patent.
- I just wasted $15,000 and years of my life.
=============================
It feels like the first-to-file system protects inventors who haul ass and get shit done, and punishes lazy (or unfortunately poor) inventors who sit on their inventions and cannot afford to /or/ refuse to move them to market.
So maybe this is the good thing?
The first-to-file provision essentially just makes it easier to decide who gets the patent if two people patent substantially the same thing. I don't think it's going to make much of a difference in the existing breakage of the patent system.