Yes. I think once you've gone through the process you should be able to rely on a presumptive validity of the patent.
The problem is with the examination process, and that needs to be fixed. I think everyone knows the presumption of validity is a bit weak in reality, because the examination process isn't as good as it needs to be.
I guess my first point of attack in solving most of the problems with the patent system would be to focus on examination. If you can actually get to a point where the only patents that are getting allowed are the ones that should be, you take care of a lot of other issues in the process.
The reason they'd be at at disadvantage is because under the treaties you can rely on activities in the U.S. (such as a filing) to establish priority in foreign countries. If the U.S. unilaterally dissolved its patent system and exited the treaty, then businesses in countries all over the world would be able to rely on their domestic activities when going for patents in other countries (EU, China, Australia, Japan, India, or whatever), and U.S. countries would not. It wouldn't hurt the big, wealthy companies as much, for for smaller entities, being able to rely on activities in the U.S. and having that extended time period to go into foreign countries can be critical.
But I think we're both in agreement that big reforms are needed. While we're at it, I'd knock the copyright term down to something reasonable as well, but that's another thread :)
Yeah, you could do that, although to be honest, I think that given the time and expense of getting a patent, and the fact that you've gone through the examination process, there is a good argument that the patent should be presumed valid (i.e. that the patent office did it's job correctly). That's in a perfect world, though, and the USPTO is far from perfect. I'm not sure the presumption of validity gets you are far these days as it used to, though. I think courts and the PTO both recognize that while you should be able to rely on the quality of an examination, it's basically a fiction at this point.
Reform, maybe, if we can muster the political will. I don't see abolition any time soon. There are international treaties involved and no incentive for any one country to unilaterally withdraw and abandon a patent system, putting their own businesses at a disadvantage in the other countries that keep them. So you'd have to get all of the most economically important countries to get rid of patents together. Seems like a long way off, if ever.
On the other hand, we have our representatives in Congress that are supposed to handle this stuff, and that's where people should be focusing attention. I don't think the Supreme Court is going to eliminate software patents with a decision. Software has been patentable for some time in the U.S., and the Supreme Court would probably look at the fact that Congress hasn't acted as indicating a correct interpretation of statute. In fact, 35 U.S.C. 101, which define patentable subject matter, has been interpreted extremely broadly since a 1980 Supreme Court decision. So the Congress has had more than 30 years to do something about it and hasn't acted. They could eliminate software patents overnight, or place new restrictions on them, or whatever. And if we want to get rid of them, that's what Congress is supposed to do. I suspect the Supreme Court will take the view that unless Congress acts to eliminate them, there is no reason to disallow them.
Yes. Or at least, a problem with how prior art is searched and applied. I guess the big question is - how do you fix that? The sheer volume of prior art references is only going to go up. The patent examinations have to be done by a person who can't consider anywhere near the total amount of art out there, and if they can't find prior art to knock out a patent application they don't have much choice about letting it through.
I don't know how you fix that completely. It could certainly be improved. Some people might argue to just get rid of patents, but that's not a practical solution because there's no serious chance of that happening any time soon. So the big question, to me, is if we're going to have a patent system how do we deal with ever-growing numbers of applications and massive increases in the prior art, when you have a finite number of people sitting in the patent office trying to examine patent applications?
Yeah. That gets back to the searching problems. Some examiner's are just bad. But most of them aren't. The problem is, the only way they can decide whether one is deserved is by looking at the prior art in front of them and comparing it to what the applicant is claiming for patent. If they're missing important prior art, then the patent may get through. Part of the reason we have the kind of proceedings that Apple is involved in right now is because third parties often have a better incentive and better resources to find the best prior art out there. So, in a way, the system is set up to acknowledge that patents are going to get through that shouldn't, and the idea is that some of the bad ones will be caught in post-grant proceedings, or in court, or whatever.
Yeah, that does make searches harder, in all art areas though it seems especially problematic in software. In theory, vague or "indefinite" patent claims aren't supposed to get through the patent office either. They're supposed to be clear enough so that someone looking at it would have a reasonable understanding of whether they're within the claim scope. That's another area when bad examination causes problems, I suppose.
With respect to software, it's probably a good area for Congress to step into, but they don't seem to be willing to do it. So we'll stumble around a bit more while the courts try to figure it out.
I think my post was in plain English, but let me summarize it briefly for one who apparently can't make it out. There is nothing in my posts regarding software patents or whether they are good or bad. My only comment was directed to why patent Examiner's, in any art area, might miss prior art. I'm assuming you didn't follow any of that, since you went off on a tangent. Maybe you're just trolling. In any event, if you want to flail around further with whatever your mind invents upon reading a post, I doubt I'll be bothered to respond again. I also doubt you're a programmer, since most programmers I know pay better attention to detail than you've evidenced here.
Usually, if the patent falls, licensing deals are going to go down with it. There may be exceptions, depending on what the license says (though if there is even a question, I wouldn't want to be in the licensor's shoes).
I don't think Samsung is likely to get any of its legal expenses back. Have they paid any of the judgment at this point?
I've seen Examiner's do Google searches. It is not the best searching method. Some of the companies that do prior art searches have better electronic systems. Google and other web search engines run into the same issue of returning far more results than the patent examiner can go through.
No, Juan. That's why I said the term "evergreening" is really misleading. They can't recover the original subject matter, and from what I know of patents in India, that holds true there just like it does in the U.S.
Instead, when companies patent a trivial modification to a drug, they use the power of their branding, marketing, distribution, and so on to push that new version to prominence. The generic company can produce the original drug, but now you have an incentive for the doctors not to prescribe the "old" version and to go with the "improved" version instead, and on the improved version there is no generic competition.
I think the reason things get caught on re-exam is that the patent office has new art before it that it didn't catch the first time around. Sometimes, this is just the result of a bad examination the first time. Sometimes, it isn't. If you think of the vast scope of prior art out there (we're at what, over 8 million issued patents, who knows how many published applications, and literally uncounted masses of non-patent literature). There's no way that it is possible for the patent office to have reviewed all of the art that is out there before making a determination.
What we need is a better prior art search process that is able to focus the search and is more likely to pull up the relevant art. That's not as easy to do as it sounds, though, because given the bulk of prior art out there, even well-constructed electronic searches of the art are going to miss references that use different wording to describe the subject matter.
The patent office does let a lot of patents through that should never have made it (the one that was recently shot down on a mathematical formula should have been unpatentable subject matter from the outset), but sometimes they do a decent job with the art that is before them, and once the patentee tries to enforce the patent, the accused infringer spends a lot more time and money on searching and comes up with art the examiner didn't have.
Yes, they could. The pharma corps push doctors to prescribe the new version. Also, doctors are sometime more likely to do it for liability reasons. If the patient dies, they don't want to be sued for malpractice and asked why they were using the older version when there's a "new and improved version" on the market.
Yeah, but that wouldn't matter. If you eliminate elements from a patent claim, you don't infringe. So if a patent claim covers A, B, C, & D, you can make A, B, & C all you want without infringing. It's when you add D that you're in trouble.
Nevertheless, I think it is a good idea that courts look at these kind of incremental 'innovations' more closely to determine whether something truly inventive is going on or not. If not, no patent should be allowed.
Even if an "evergreen" patent goes through, can't the public still practice the claims of the original patent?
In other words, if a drug company patents a medicine having components A, B, and C, and then when that patent expires they get another patent for the medicine, but having components A, B, and C' (where C' is some modified version of C), then generic drug companies should still be able to produce a medicine having A, B, and C without infringing the new patent.
If you're talking about a private employer, the issues are not the same because there are no Constitutional protections at work. Remember, the Constitution constrains government action. If a public school district, or public official, or public employer does this sort of things, the First and Fourth Amendments come into play. If a private individual or employer does, the Constitution is not implicated and you have to rely on statutory or common law (to the extent it covers the act).
I wonder why they don't mention "obviousness?" It is true that for a patent claim to be anticipated, each and every element of the claim has to be present in a single reference. However, to be valid a patent claim must also be non-obvious, and that requirement of a single reference doesn't exist in that case.