Slashdot Mirror
Byte Offers An Explanation Of Patent Law
dubl-u writes "I just ran across this detailed explanation of patent law in a recent Byte column. It made a lot of things clearer for me. There's one great tidbit: there's a special kind of cheap patent that's perfect for open-source projects!" You will have to hit that link to find out what he's talking about, right?
I found this exerpt to be of particular interest:
Public Law 106-113 introduces a new "inter partes" patent reexamination procedure that lets any third-party initiate an interactive reexamination of a patent based on new evidence of prior art. By facilitating the two-way exchange of argument and evidence, the PTO hopes to mediate many of the disputes that currently end up in legal action. When a reexamination leads to a finding that patent claims or portions of patent claims are invalid due to the existence of legitimate prior art, the claims can be stricken or revised in reasonable ways and a "reissue patent" will be published. A patent may even be declared invalid in its entirety as a result of reexamination. Providing this inter partes reexamination mechanism should avoid a lot of unnecessary lawsuits and keep the PTO squarely in the middle, and therefore accountable for its mistakes, when infringement or interference disputes arise.
Plenty of projects, not enough developers...
They have always been a fantastic source for unbiased and acurate reporting on computer trends from the hobby days of computing up to present day world-wide distrubution on the web.
Any platform, any language, any diciplin, Byte has always been an interesting and informative read for many many people in the IT field. I miss the articles, and I miss the unbiased insights. I still have my "wall-o-Byte" archive on the bottom row of my bookshelf and would like to continue adding more print editions in the future. What does it take to publish a magazine these days?
_________________________
Provisional patents solve the smaller of the two giant problems out there. That is, they solve for the laziness of the patent office in doing research.
However, it only solves for it spottily and does not solve the problem of people getting silly patents that they can then use to shut down open source efforts that might hurt their proprietary software sales.
The right way to deal with this would seem to be to establish (as has been discussed elsewhere) a non-profit "patent custodian" that people can donate their patents to. Then have that organization use licenses for its patents as a lever to extract open source licensing for other key patents. For example, if the GNOME folks had a few good patents, they might be able to get Adobe to release their color model handling for use in GNOME programs (and their derivatives). This really should not be an issue, since any GNOME derived program is subject to the GPL, and that means that most of Adobe's rivals won't touch it.
good examples for some of the projects out there are the way Perl does its advanced regular expression stuff and the way that Slash manages moderated discussion forums. If there were patents on all of these things, licensed to any open source effort that wanted to use them, the patent situation might be livable.
I don't think this will happen, though, until the open source world begins to truely believe that the USPTO will not be magically reformed and turned into an organization with a clue.
I remember my dad's stacks of Byte magazines when I was growing up.
Last time I was home, i dug through some.
10 Mb MFM Hard Disk for your PC-XT.... $2500 bucks!
It appears that this type of patent does exactly what Jeff Bezos wanted to do when he patented one click technology -- it establishes prior art and prevents someone else from coming along and patenting the same technology. If Jeff Bezos was telling the truth when he said Amazon was using the patent system as a defense mechanism to prevent future competitors from copying their technology and putting them out of business, the SIR patents would seem to be the way to go. It wouldn't give Amazon the right to sue anyone for damages but it would put a stop to anyone from stealing their "technological innovations".
I wonder how Jeff will talk himself out of this.
Slashdot...thanks for the great story. If we could moderate stories (which I think is a good idea), I would give this a +5 for informative. Keep them coming like this !
Hates people who have stupid little sigs
Actually this won't do squat to help us. It only protects us in those instances where we created and patented the specific invention in question before anybody else.
:).
Unfortunately these overly broad and near meaningless Internet patents, Business module patents and software patents that are stirring up trouble would still be a problem. How many OSS diehards would have thought of patenting "One click shopping" 2 years ago? I am willing to bet that even the geeks at Amazon who implemented this code didn't think of patenting it until the patent lawyers Bezos higherd went looking through the lab for anything add.
However in the event that we did have a cheap Patent on "One click shopping" we could still be sued for abusing the "Affiliates patent". In that event we would have no defense but to search for prior art to the Affiliate program. Just as if we had no patents at all.
However if we had a full patent on something ridicules like "threaded discussion on a web page" and someone decided to sue us for violating some unrelated patent ( Like "One click" ) we would only have to find evidence of them using Zope's "SquishDot" and the case is over. Out of court settlement with a written agreement to never file patent suite against any free software developer.
This is where Amazon can show weather it's a jungle or just a big river. Put the ridicules patents it owns into a pool managed by the FSF and LI. Ask every member of LI to contribute some patents to the pool. Then we can come back and talk about being attacked. Simply put it's not that difficult to assemble a set of 50 to 100 ridicules patents that every new or dynamic business on the planet violates one of. ( Those that have done nothing new for 30 years are protected since everything they do is prior art
--= Isn't it surprising how badly I spell ?
Well personally I would rather abolish patents
alltogether...however...perhaps a slightly more
complex system:
Certainly I think software patents are just plain
wrong. Especially ones dealing with encryption
(you could easily break some patents with a
pencil and paper).
Here are a few things that would make patents
much less diagreeable:
A) Applicant must show intent to actually use
the patent (ie actually produce what is being
patented or use it in a real product). Alteratly,
perhaps an applicant may be granted a temporary
(1 year) patent which will become a real patent
after the current holder (applicant or someone it
was sold to) shows intent to use it.
(sueing people alone should NOT count as "use")
B) Patents should be granted for "an apropriate
amount of time". It should be determined by what
it is and what feild it is for. example: when IBM
applies for a hard drive technology patent...how
long after the patent is filed, on average, does
it take before they can actually deploy it in
a real product? A 3 year patent isn't very
good if it takes 2 years to actually finish and
get a woking product out the door.
C) After the USPTO has made a preliminary reuling
to allow a patent...for a period of 3-6 months
the patent should be made publically
avilable...and a period should begin where the
public are allowed to submit prior art etc
for consideration.
D) Submarine patents. If a company can show that
they were using a technology before the patent
was actually issued (but after its development
by the patent holder - ie they didn't read the
application or copy the idea) then they should
NOT be fully responsible. In fact...the patent
holder should be forced to give them a compulsory
licence to the technology (ie he can't refuse it
and shut them down) for a reasonable royalty, of
FUTURE moneies (ie he should have no claim to
the money from before the infringer knew about
the patent)
[btw the idea of a compulsory licence does exist
in copyright law...you can distribute copies of
music without obtaining any permission legally...
but you MUST follow specific procedure and pay
royalties]
Just my thoughts....any others?
-Steve
"I opened my eyes, and everything went dark again"
I'll reiterate, you're very naive.
There are a couple areas of industry that really need the protection that patents give. These are industries such as pharmeceuticals, where the amount of time and money put into developing a new drug is so enormous that the patent serves as a way to earn back those expenses before your competitor can manufacture it and sell it for less. Pharmeceuticals are also the reason why patents currently last so long. Since the patent is usually applied for before testing begins, by the time lab testing, animal testing, limited human trials, larger clinical studies, and full FDA approval are granted, there may only be a couple of years left on the patent. And, up to this point, the drug company has invested millions of dollars with no income.
Without patent protection, you can kiss any new drug research goodbye.
That being said, there needs to be much more differentiation in the amount of time a patent lasts. For pharmeceuticals, 17-20 years is fine, it's not a lifetime monopoly, since many drugs stay on the market for 50 years or more, after release.
For "business models", algorithms, software, and other non-tangible products, patents should only be granted for 1-3 years. Patents were intended to allow a new business to recuperate R&D costs, and get a foothold in the market before their competitors could use their ideas. They were not intended to give full life-cycle monopolies.
The system is broken, but it's not evil.
- A patent applicant must show intent to actually use the patent. For utility patents, this means marketing and selling the device, other patents work as appropriate. No intent to use, no patent.
- A patent expires if the patent is not "used" as defined above within one year of its filing. However, an applicant can get an extension on this by showing evidence that the device could not possibly have been marketed within that time. Research reports would suffice (showing that the product is still in the development stage). The patent can be extended as long as the applicant can continue to show that the patent cannot yet be marketed, up to a maximum of five years (two years for software-related patents).
- A patent expires at a set amount of time after the patent is first "used" (as defined above). This time period is the same seventeen years for individuals, five years for corporations (three if the patent is software-related). If the patent transfers, the time period is adjusted accordingly (so if a person holds a patent for six years and sells it to a corporation, the patent immediately expires). This amount of time cannot be extended by any means.
No doubt the most controversial part of this will be the different time periods for individuals and corporations. The rationale behind it is that a corporation has significantly more resources than a person does. One person would be lucky to have more than one truly significant invention in his lifetime (truly major breakthroughs -as in, on a scale you very seldom see from individuals or even corporations- notwithstanding). A corporation which can't push out something new and patentable in five years needs a new R&D department.There are a couple areas of industry that really need the protection that patents give. These are industries such as pharmeceuticals, where the amount of time and money put into developing a new drug is so enormous that the patent serves as a way to earn back those expenses before your competitor can manufacture it and sell it for less. Pharmeceuticals are also the reason why patents currently last so long. Since the patent is usually applied for before testing begins, by the time lab testing, animal testing, limited human trials, larger clinical studies, and full FDA approval are granted, there may only be a couple of years left on the patent. And, up to this point, the drug company has invested millions of dollars with no income.
Yes, I feel so sorry for the huge drug companies that are using their patents to gouge consumers charging ridiculous prices for drugs when they could easily charge less and still turn a profit. As for the investments they make, american drug companies are highly subsidized by the United States government, sometimes paying millions of dollars in development costs while the drug company reaps all the benefits. Furthermore, the drug companies have effectively used patents and WTO procedures to block companies in third world nations from making cheap versions of their drugs that could be used to save thousands of lives. Instead, they must rely on expensive drugs made by american companies which of course they cannot afford and thousands die.
Thank you and good day.
Yet again, we are saved by patent laws that some feel are evil, and should be done away with.
;), and you can compete with the nasty ugly corporations(of which we hope you have not become).
Think about it this way: without these laws, the rich would be richer, and the poor would be dead. You, a fairly capable inventor/developer, have created a nifty gadget that would sell fairly well. You start manufacturing, and then a rich guy down the street copies you. He has slave labour in a third world country somewhere, he makes them cheap, and sells them cheap. All of a sudden, the life savings you spent making this thing are down the tube. And there's nothing you can do about it.
Now, imagine a world with intelligent intellectual property patent laws. You, the inventor, patent your invention. You have twenty years(or something like it), to make your money. The patent expires, and everyone starts making your gadgets. Luckily, though, you've had time to ramp up production(hopefully not using slave labour
I once read someone say that without very strict patent laws, open source would be impossible. And I agree. The GPL is based on those laws, and others like it. If they didn't exist, there would be nothing stopping people from using code developed in a free and encouraging community in their own proprietary works, without recognition, without compensation.
The SIR patents seem to be a windfall to the open source community. With these and the GPL behind us, we can lay the groundwork for a re-invention of computers. The way we use them, and how we understand them.
I am not lawyer, nor am I a rich man. But, if someone set up a charity drive/fund for open source SIR patents, I would gladly donate what I could spare.
David B. Harris
Barclay family motto:
Aut agere aut mori.
(Either action or death.)
Why do patents get examined at all?
As I understood it, the idea of patents being examined by the patent office was to determine if they could be valid -- in other words, to determine if there was prior art, and if the invention was sufficiently 'novel'.
Why do this at all?
Wouldn't it be perfectly good enough to simply have a 'registry of neat ideas', where people could, for some tiny fee, register a neat idea they had, after which point anyone wanting to use the idea would either have to negotiate with the inventor OR show prior art OR show that the invention is not 'novel'?
It seems to me that this is effectively what happens already -- the patent office seems to issue patents on basically anything, and the validity of the patents is only going to be sorted out in the courts anyway.
Tarsnap: Online backups for the truly paranoid
If we did set up an organisation to collect and hold patents for the Open Source community, we could use:
;-)
http://www.opensirs.org
(Say it out loud)
Gerv
Instead, they must rely on expensive drugs made by american companies which of course they cannot afford and thousands die.
Here you're assuming that the drugs in question would exist at all otherwise. There's ample evidence that that isn't the case. In fact, the US system of patents and incentives has clearly been responsible for drugs that treat Crohn's disease, leprosy, Sickle cell anemia, particular types of lymphomia, Tourette syndrome and cystic fibrosis, as well as a lot of the development of the fast-growing genomics industry--an industry which has the opportunity to really change medicine.
In the end, often this comes down to one of the most difficult kinds of moral choices: Do I save 1,000 people today, or 100,000 in ten years. Dump patent protections and you may very well save 1,000 people today, but you may condemn thousands or millions to death in the future who could have been saved by new research. The question is made no easier by the uncertainty in the results of future research, but the trends in genomics are very encouraging.
A real argument can be made for either answer, but make no mistake, the real moral question behind this issue is not so simple as "drug companies kill thousands."
--j
I'm a nature photographer.
here's one great tidbit: there's a special kind of cheap patent that's perfect for open-source projects!"
Wrong on both counts. SIRs are not patents and they are not cheap. A "Statutory Information Registration" permits an applicant who has no chance of getting a patent issued, to have the USPTO print out a pretty certificate with his disclosure, complete with an issue date, a number and everything. One doesn't need to file an application, you can ask for a SIR from the get-go. (But there is no real reason to do so.)
It is not a patent, does not give you the right to place a patent number (or patent pending) on a product and grants no right to exclude. While the SIR is not examined against prior art, it is examined for compliance with all the formalities and disclosure requirements of Section 112 of the Patent Act, and thus should be reviewed by someone experienced in preparing patent applications.
Ultimately, it has no legal effect at all, except that it serves as prior art to the extent of its disclosure. On the other hand, so does every other publication, whether issued by the PTO or not. We would be far better off publishing a "Journal of Open Source Invention," to provide a source of prior art than relying upon SIRs.
And they are certainly not cheap. The patent office fee for an SIR is $920.00. (As compared to $620.00 for a full U.S. Utility application filed by an individual).
For the open source community, I see no upside for SIR applications.
There is a more informed discussion of this issue over on technocrat.net .
There is a HUGE problem with the new inter partes reexam provisions. If the patentee prevails before the PTO in the reexam, the third party CANNOT appeal the decision to the Court of Appeals for the Federal Circuit (the exclusive appeals court for all patent cases) and cannot assert the same piece of prior art as a defense in a later infringement lawsuit. However, if the third party prevails, the patentee CAN appeal. This is a tremendous downside to the process because the third party is forfeiting its day in court in favor of a decision by the PTO but the patentee faces no such risk.
Provisional patents solve the smaller of the two giant problems out there. That is, they solve for the laziness of the patent office in doing research.
There is no such thing as a provisional patent. You can file a provisional application for a utility patent, which can be "re-upped" into a full-fledged application if you did it right. But nothing here would have the results you claim.
If you are referring to Statutory Information Registrations, referred to in the article incorrectly as "SIR Patents", they are a joke. See my earlier response on that subject.
It appears that this type of patent does exactly what Jeff Bezos wanted to do when he patented one click technology -- it establishes prior art and prevents someone else from coming along and patenting the same technology.
As to prior art preclusive effect, a SIR is no different from any published article. And the mere existence of prior art doesn't mean that a given application will have the SIR cited against it. (As noted, SIR's are more expensive in terms of filing fees than issued patents for most entities).
I am not lawyer, nor am I a rich man. But, if someone set up a charity drive/fund for open source SIR patents, I would gladly donate what I could spare.
You would be wasting, or at least inefficiently using, your money. SIRs are not cheap and they are not patents. For far less money than the minimum $920 per disclosure, we could establish a foundation to publish (and provide quality re-editing) an open source journal for these processes, which would have equally effective prior art effect.
It would be far cheaper, and far more helpful for the movement, if the same applications (to the extent patentable) were filed as applications, matured into patents, and pooled for use in cross-licensing against the real bears. This presumes, of course, that we are actually creating patentable inventions out here in the open source community.
The perception that the office issues patents on "basically anything," while understandable, is greatly exaggerated.
It is certainly true that a decent lawyer can probably get something out of any disclosure, it remains to be seen if the claims allowed have any meaningful effect. In most art areas, a survey of USPTO records provides a fairly complete account of the state of the art. Unfortunately, this is not so in software arts and methods of doing business, owing largely to the history of decades of practice (centuries in the latter case) before patents were permitted.
There already exist "registries of neat ideas." They are called trade and academic journals. And they already constitute prior art.
I have a proposal which I think more effectively balances the relevant considerations more adequately than the status quo. You will also find there a brief discussion of the examination process, what it does and more importantly what it does not do. But I certainly would not do away with examination.
Trust me, its not as easy in practice as it may seem.
Umm.. no. Did you even read what I wrote? Patents should be given only to individuals not representing a corporation.
Your proposal has a lot of very severe problems. For example, how are you going to get corporations to fund research into new technologies when they don't end up owning the rights to the technology. If you actually are succesful in preventing corporations from owning patents, it would be the end of commercial R&D in the US, which would result in much slower rates of technological progress - and economic growth. Look at the recent economic expansion we have had - the Federal Reserve believes that much of the credit for the combination of low unemployment and low inflation is ddue to increasing worker productivity fueled by technological change. Do you really want to throw that in the crapper?
Even under current law, patents are granted to individuals (the inventors) now. The way corporations end up owning the patents is that the inventors are working under standard employment contracts that require them to assign the rights to the patents to their employers. If you try eliminationg this, the employer will just change the provisions of the employment contract to something like requiring a royalty-free exclusive license from the employee.
The prohibition against appeal or use in a future suit by a third party is explicit in the law itself. Find the law at
1 948gb1.pdf
http://www.uspto.gov/web/offices/com/speeches/s
Appeal: page 49 of pdf
Future suit: page 44-45 of pdf
> Well, probably A is covered by paying upwards of
> $20k over 2-3 years - if that's
> not a form intent, then what is?
What I had in mind, maybe I didn't specify well.
I think that if patents exist then anyone with
a patentable whatever should be able to get a
patent...so $20/k for 2-3 years sounds like a
very high "bar" to get over.
What I would like to see is that if a company has
no intention to use a patent for something other
than legal bullying, then they should either
not get the patent, or lose their patent.
> B) Right now it appears 17 or 20 years is
> considered the appropriate amount of time.
I personally think it is too long. Of course
perhaps 17 years would be a good maximum if
A was implimented (ie...a company shouldn't
be able to stop producing a product, and then
use its patent power to force others to pay up.
if the current holder has no intention to use
the patent for other than bullying...they should
lose it)
C) I did not know that. Very good to know.
> D) - Isn't this just commercial good sense?
In some way. Again,...the goal of my idea was to
think of a system where a patent does what it
needs to, and doesn't function as a club to
beat people around with. \
of course as I said...I dislike patents anyway.
I generally think that recognizing a good idea,
then taking it and applying it where you need
it is a "Good thing" and patents seem like a
device to get in the way of people using good
ideas.
"I opened my eyes, and everything went dark again"
There's a registry for non-patent prior art at the Software Patent Institute. The USPTO searches that database, which contains hundreds of thousands of old manuals, papers, and such. The emphasis is on "old"; though; there are over 1000 UNIVAC references, but only two Linux references. It's useful, though, in that most of the classic algorithms are in there somewhere.
It's well known that the USPTO is granting patents which never should have been granted, patents on something which is already well-known. For example, Unisys managed to patent a simple linked list, and the LZW algorithm is patented _twice_ - the USPTO couldn't even manage to check the later application against previous patents.
Now the patent office uses software, right? Here's what you need to do, if you have the money:
- Apply for a broad patent which will cover some software used by the USPTO. This software could even be Windows, sendmail or whatever.
- Hope it gets granted. If not, try again with a slightly different application.
- Sue the USPTO for patent infringement and refuse to license the patent to them under any conditions.
- With luck, the patent office will be forced to go to court arguing that the patent should never have been granted.
- Hope that newspapers pick up the story.
-- Ed Avis ed@membled.com
So, basically, an SIR is an OpenPatent? It is a patent which servers /only/ to relinquish all rights to the public domain, and to prevent others from claiming a similar patent. That's good news. I suggest all programmers and companies (in the software industry) start using these "ethical" patents.
It's 10 PM. Do you know if you're un-American?
"Why? It simply costs a lot of money and doesn't theoretically gain you anything over the cheaper but just as ethical method of simply publishing the idea."
/find/ your publishing to establish prior art. From the recent article on the patent office, we know that patent office workers don't spend near as much time as they should researching prior art. I'm not very familiar with the Open Patent License, but if it does a better job than an SIR, then great. I just wasn't aware that there was already sort of a provision for this sort of thing.
Because merely publishing the idea is a lot less effective in keeping others from patenting that idea. Merely publishing the idea puts the responsibility on patent office workers to do due diligence and actually
It's 10 PM. Do you know if you're un-American?