Slashdot Mirror
EFF Announces New Patent Reform Project
olsmeister writes "On Tuesday, the Electronic Frontier Foundation announced its 'Defend Innovation' project, which includes seven proposals for software patent reform. These proposals include things like shorter coverage for software patents, and a requirement to demonstrate running code for each claim in the patent."
extreme fines and penalties, including loss of some existing patents, as punishment for not mentioning 'prior art'.
In other news, the EFF has adopted Don Quixote as their Official Mascot.
That's cute. I wonder whether anyone will listen.
No more granting patents for the bloody obvious! Things like Amazon's one-click, or Apple's patent for a slide-to-unlock bar. This includes things which may sound arcane and innovative but are actually obvious to anyone working in the field, which in turn means the examiners will need to be better trained.
Or how about a 'fail fee' to discourage companies from fileing thousands of junk patents in the hope that just a few will be approved by luck and can be used to fill their patent chest?
A ban on patents that provide no obvious benefit to mankind, but rather are intended specifically to prevent interoperability with a specific format? Things like filesystems or file formats. I'm thinking specifically of MSs patent on long file names in fat32.
The only way to fix the patent problem is to shove GOBS OF MONEY down the throats of ever hungry politicians and their banks.
In a day when even the courts are questioning the validity of the very concept of software patents, EFF should be taking up this issue. Instead, they are lobbying for "shorter durations"???
Boo, hiss. EFF, I expect a lot more from you guys.
That's cute. I wonder whether anyone will listen.
My cynical side says, "Not unless they have as much money pumped into lobbyists as *IAA does". My realistic side says, "Not unless they have as much money pumped into lobbyists as *IAA does"." Oh... damn.
I think what might be interesting is to develop a database of prior art/ideas. With thousands of people submitting everything they can think of every conceviable obvious aggregation of technologies any patent application would be required to be checked against the database and rejected if someone else already dreamed of it before the application was filed.
The database would be run by the patent office or it would incorporate cryptographic and distributed features allowing timestamps and content to stand up to any challenge.
Companies not wishing to pay the patent tax might seek to use it in a bid to prevent others from filing first and locking up the same blatently obvious ideas for the next 20 years.
I want a law that: - Force that all source code of commercial software products must be allowed for inspection of the customer (source code under the license the manufacturer wants). - Allows abandomware to be use legally. - Force the release of abamdomware (discontinued software) source code under an open source license.
But ultimately this is still pro-software patent, they just try to modify the existing rules somewhat.
I have seven proposals:
1. Get rid of all patents.
2. Get rid of all patents.
3. Get rid of all patents.
4. Get rid of all patents.
5. Get rid of all patents.
6. Get rid of all patents.
7. Get rid of all patents.
After the first one, the rest are easy. :)
So a big company with the HR resources to fast-track an implementation of an idea it rips off an individual inventor would have the advantage...
"Politicians and diapers must be changed often, and for the same reason."
Do you all live in the real world or has the dark of the basement really cut you off from outside too long? If they outright try to lobby for radical change to the system it won't get passed and the naysayers will be emboldened and any hopes of change will be harder to accomplish. At least the last point, research whether they help in the first place, sets groundwork for proving whether they help.
FFS, lasting revolution doesn't happen overnight, sometimes the wiser thing is to change things slowly and smartly.
If the idea doesn't work and cannot be shown, it's just that, an idea. 10 a penny fodder.. A rip-off artist would have to know about the idea before patenting, duhhh! If an implementation had to exist and work before patent submission, the rip-off artist would be limited to poaching the staff or buying the company out.
I've said before, in other threads, that I'm a patent attorney. I have my undergrad degree in CS. I have an ill view of many patents that are litigated. That's not to say there's no innovation happening in the software world, or that there's nothing worthy of protection.
Nevertheless, I generally think that the EFF's proposals here are misguided. Taking them one by one:
Part of the problem in the first place is that it some times takes longer than 5 years to even get a patent. Most of the term of a patent (under this scheme) would be consumed with prosecution. It doesn't account for the Patent Office.
On board with this one, but it doesn't go far enough.
We really should be talking about fee shifting more generally in the case of a patent owner that fails to do their own diligence, but rather is seeking to extract nuisance money or impede lawful commercial activity.
This doesn't address the real problem. The real problem is when old claims are read to cover future technologies that they didn't not have possession of at the time the application is filed.
In a non-software example, it would be like a patent claim that covers something no larger than "a micrometer" before "nanometer" manipulation was available.
There is already a prior user defense baked into the most recent patent reform bill.
Patents are. Most patent applications are public within 18 months of filing too.
I'm not sure why a license must be public. That's just a contract between two people.
I'm also not sure what public records they're referring to. It could be that patent assignment records aren't "up to date." (http://assignments.uspto.gov/assignments/?db=pat). I could get on board with requiring those be up to date.
That's all relative. Besides, plaintiffs still need to prove damages. Defendants still have an opportunity to rebut those damages.
I bet you can predict how that'll turn out. In any case, doesn't move the needle much.
If the idea doesn't work and someone gets a patent on it, no-harm, no-foul, right? Nobody is going to infringe...
If the idea does work and it has yet to be shown to work in a product, a large company might have an advantage, right?
Sometimes smart folks think alike. This is why companies usually want to start patent filings as early as possible, usually well before the idea can demonstrate actual advantages (e.g., usually as soon as partially functioning prototype or even half baked pseudo-code/flowchart suggests it can be refined into potential advantage). Then they stretch out the filing by adding refinements to their patent claims as the implementation gets fleshed out and patent goes through the review process.
Invalidating the patent? That should have them quaking in their boots. Kind of like if you rob a bank and get caught, you have to give the money back. Does explain why a certain patent application went through even after being notified of prior art.
IMO the problem with software patents is that there is no way to know if your software infringes one.
Only #4 "Infringers should avoid liability if they independently arrive at the patented invention." even comes close to addressing the problem.
My suggestion is this;
Any computer built more than 1 year prior to the filing date of a patent, does not infringe that patent, regardless of how it is programmed.
Likewise any software that runs on a computer built more than 1 year prior to the filing date of a patent, does not infringe that patent.
Note that a strong argument can be made this is already true. I think we should lobby for a law that spells it out explicitly.
As I understand it, patents were originally supposed to cover specific implementations, not general ideas. You had to have a working physical model. This has gradually expanded until patents do in effect cover ideas. Patents merely list every possible way it can be implemented, as broadly as possible.
Now algorithms, software, and so called business methods can be patented. Patents can be effectively renewed by tweaking the ideas and applying for a new patent. There are lots of other tricks. They all make life difficult for people who want to concentrate on innovations, not arcane legalisms.
Why did we end up going this direction? We didn't have to turn patents into the oppressive, strangling, chilling monstrosities they are today. They were supposed to help the little guy, but were too readily made into tools that large organizations could wield to help maintain their dominance. True, they have to fight off the occasional troll, but they evidently think the control they gain is worth that.
The entire approach of patent law is all wrong. It's all about control and preventing loss and "theft", and security against our fears no matter how worthless the former and ridiculous the latter, when it should be about sharing and gain. You should ask for permission first? No one does that! Permission can be flat denied, holders don't have to grant permission for some standard price. One way a business can handle this mess is to build up a defensive patent war chest. Fight fire with fire. Cross license. It's a lousy way that perpetuates lawsuits and enriches lawyers. So I think the system ought to be radically reformed, or just plain scrapped. A patent should never be a means of denial, of retarding innovation, and squelching competition. That's the very opposite of their intent. Dump the monopoly part at the least.
Intellectual Property is a monopolistic, selfish, and defective concept. It is "tyranny over the mind of man"
I have long argued in favor of reducing the term for software patents. Although a 20-year term makes sense when applied to, for instance, pharmaceutical patents (which need at least a couple of decades in order to recoup R&D expenses), it makes very little sense for software patents. The patent system needs more flexibility so that such a multi-tiered arrangement could be pun in place; the one-size-fits-all approach is no longer the most effective model.