|
I'm not sure how that's different than my explanation but with a game theory reference. I have given several reasons - you either haven't responded to them or basically told me I didn't know how the industry I work within works  E: to be fair, I suppose I was restating OneEightHundred's argument, but still: he's right E2: I'll also grant that I've probably done a poor job articulating my arguments because I don't have to on a normal day. "Patents" (or some form of the word) is a universal gallows humor punchline among the people I work with because all of the cases we hear about are so goddamn stupid and frustrating and the patents we see (and produce!) are total loving nonsense that we all just see the system as obviously broken garbage that we'd all just as soon scrap as try to reform. Munkeymon fucked around with this message at 23:56 on Dec 16, 2013 |
#
?
Dec 16, 2013 23:07
|
|
|
|
| # ? May 14, 2024 12:33 |
|
|
It's almost like the only ones that make the news are the ones that upset people or are borderline. Do you think that there might be a selection bias operating?
|
|
#
?
Dec 17, 2013 02:00
|
|
|
Munkeymon posted:I'm not sure how that's different than my explanation but with a game theory reference. The game theory explanation is very different than stealing someone's marketing research. That explanation suggests that if the first chain to open a coffee shop gets their marketing research wrong and puts it in the wrong place, the second company (even if it knows the marketing research is incorrect) should still place its store next door. I'd agree with what Kalman said about selection bias, but I don't think I accused you of not knowing how your industry works. What about process patents generally? How do you feel about the PCR patent or transgenic organism patents? Corn can be replicated almost as easily as code.
|
|
#
?
Dec 17, 2013 02:34
|
|
|
Kalman posted:It's almost like the only ones that make the news are the ones that upset people or are borderline. Do you think that there might be a selection bias operating? Not every patent needs to be frivolous to cause serious impacts on innovation. In terms of lawsuits, patent-assertion entities make up 60% of patent lawsuits filed in 2013. Not every patent from a PAE is frivolous, but the number overall is probably fairly high. quote:Patent-assertion entities, also known as P.A.E.’s, typically have no operations other than collecting royalties on patents. They accounted for more than 60 percent of the roughly 4,000 patent lawsuits filed last year, up from 29 percent two years earlier. The chance of a patent being revoked once it's opposed is roughly 50%, and in most of the rest of the cases it's significantly narrowed. Obviously frivolous patents are more likely to be opposed than high-quality patents, but it trends towards the same rough numbers. My gut instinct here is that the percentage of frivolous patents is 10-30%, and many of the patents that are not outright frivolous are probably written overbroadly. It's probably higher than average in software and technology patents because it's easy to use some fancy language to obfuscate "in a program on a computer". Paul MaudDib fucked around with this message at 03:23 on Dec 17, 2013 |
|
#
?
Dec 17, 2013 03:07
|
|
|
Kalman posted:It's almost like the only ones that make the news are the ones that upset people or are borderline. Do you think that there might be a selection bias operating? From those, the arithmetic coding patents stunted the progress of data compression in general for an extremely long time because everyone decided to avoid it by using less-optimal Huffman coding rather than deal with patent garbage. The usual option of inventing a better mousetrap was off the table because arithmetic coding is provably optimal. After over a decade, it's finally starting to get supplanted by range coding, which only started coming into use when someone found out that it predated arithmetic coding and was only slightly worse. To elaborate on how ridiculous the whole arrangement is though, range coding works with an arbitrary base, and arithmetic coding is just a range coding in base 2 rephrased via sixth-grade algebra, so they managed to stunt data compression for a decade by patenting rephrased prior art combined with the number 2. Wavelet patents have done a similarly-destructive job on preventing that technology from being adopted, and everyone is sticking with DCT instead. Video encoding is full of extremely trivial patents, several that can be expressed in under 20 characters of C++ code. Codecs are not just patented as codecs, there's pretty much a separate filing for every little bullshit improvement that took more than 5 minutes of experimentation to develop. The notion that the state of the art is being advanced by patents here seems really absurd because the advancements in those fields mostly consist of deliberate patent avoidance (i.e. Vorbis), improvements made by people in jurisdictions that allow them to ignore the patents (i.e. x264, ffmpeg), and products that are black-boxed to hell because one of the easiest ways to avoid a patent lawsuit is not say how it works (i.e. Windows Media). OneEightHundred fucked around with this message at 05:28 on Dec 17, 2013 |
|
#
?
Dec 17, 2013 05:16
|
|
|
Paul MaudDib posted:Not every patent needs to be frivolous to cause serious impacts on innovation. In terms of lawsuits, patent-assertion entities make up 60% of patent lawsuits filed in 2013. Not every patent from a PAE is frivolous, but the number overall is probably fairly high. http://www.patentlyo.com/patent/2013/03/chien-patent-trolls.html The PAE "increase in filings" over the past two years is almost entirely due to a change in litigation rules that prevented them from naming multiple unrelated defendants in a single suit, which appears to have had a follow on effect of reducing the number of small defendants being sued. A proposed reform in the most recent round of legislation (mandatory stays of customer litigation if a manufacturer offers to stand in their shoes) should pretty much kill off the most egregious tactic left, customer lawsuits. And the availability of IPR proceedings (which run from 50k-250k, roughly, compared to litigation costs) has already started to make the bigger trolls somewhat more cautious about filing on weak patents to receive nuisance value (because it's fairly easy for a company to put a patent into IPR, which comes with a near-automatic stay on any accompanying court litigation and the discovery costs that go with that, and forces the troll to spend money defending their patent - money they can't recoup which means they generally can't use contingency counsel for it.). IPRs are a new process, so we're only just starting to see their effects - we only got the first final written decision on one a month ago. Trolls will likely adapt their models, but the existing reforms will probably do a lot as is; combine them with the second wave of proposed reforms and you'll see a lot of the abusive litigation tactics go away. (Without, you know, doing stupid things like trying to make things implemented on general purpose processors unpatentable while allowing special purpose processors that do the same thing to be patentable.)
|
|
#
?
Dec 17, 2013 05:40
|
|
|
Kalman posted:It's almost like the only ones that make the news are the ones that upset people or are borderline. Do you think that there might be a selection bias operating? Yep, but I've yet to hear that retort followed up by a legitimate success story. It was either This American Life or Planet Money that did a patent show and got a supposed success story from Intellectual Ventures that turned out to be bullshit, but that lie was the closest I've ever heard to a software patent success story that might have involved a patent that might have been valid. Also the company I'm currently at was sued over a bullshit patent that was basically "resize an image before you use it in a video". It was hilariously stupid considering the script to make AVISynth do that is shorter than this sentence. Here's the first result for Googling "AVISynth show image": http://arstechnica.com/civis/viewtopic.php?f=6&t=546038 Thread full of patent violating posts right there. Well, technically we don't know that because of course we settled for an Undisclosed Sum, but that's effectively the same thing as an admission of validity to a third party wondering if they should fight it.
|
|
#
?
Dec 17, 2013 06:20
|
|
|
Just off the top of my head: Dolby vs RIM (audio compression) IBM vs Amazon (interactive online interface patents) Google v. BT (qos/capacity management) Broadcom v emulex (10gigE) What, that last one seem out of place? Read claim 20 (us patent 7058150). Can software infringe that? It needs an A/D, since there's an analog sampling step, but you could easily read it onto pure software. Do you see why I question there being any real ability to ban software patents without it being either a completely ineffective ban or taking out far more than ever intended? The proper response to weak software patents is better examination, not some attempt to say that software isn't patentable. (Also, a settlement is almost never an admission of validity and I've never seen a lawyer treat it like one. It's a "oh they settled out for nuisance value? Well, let's do our diligence and see what the troll offers us as nuisance value and if it's under the right threshold we will pay it off.")
|
|
#
?
Dec 17, 2013 06:50
|
|
|
Munkeymon posted:Yep, but I've yet to hear that retort followed up by a legitimate success story. It was either This American Life or Planet Money that did a patent show and got a supposed success story from Intellectual Ventures that turned out to be bullshit, but that lie was the closest I've ever heard to a software patent success story that might have involved a patent that might have been valid. You do hear about the success stories, but you don't register them on the patent spectrum because the intellectual property aspect of those stories is exceedingly boring. The story goes like this: Inventor comes up with a great idea.** Inventor starts a small company.*, ** Inventor files for patent on idea.* Inventor gets patent.* Inventor gets funding from venture capitalist?* Company gets bought out for millions of dollars.** * Not necessarily in this order... ** This is what is reflected in the news story about the company sale. The story for non-successful stories is as follows: Inventor comes up with a great? idea. Inventor files for patent on idea. Inventor starts a small company. Venture capital? Patent issues. Company goes bankrupt. Sells patent as asset. Purchaser of patent asserts the patent for return on investment. A lot of people will tell you that the purchaser is a patent troll.
|
|
#
?
Dec 17, 2013 07:37
|
|
|
KernelSlanders posted:The game theory explanation is very different than stealing someone's marketing research. That explanation suggests that if the first chain to open a coffee shop gets their marketing research wrong and puts it in the wrong place, the second company (even if it knows the marketing research is incorrect) should still place its store next door. Company A sees Company B doing a Thing, assumes they did research that told them that Thing was a good idea that will make them money, and copies the idea behind that Thing. Company B still did Thing knowing that Company A would see and react to it - possibly by copying the idea behind Thing - because Company B still thinks it'll make them some money. That's what I was talking about, but whatever - it's a pointless derail. quote:I'd agree with what Kalman said about selection bias, but I don't think I accused you of not knowing how your industry works. What about process patents generally? How do you feel about the PCR patent or transgenic organism patents? Corn can be replicated almost as easily as code. You think companies are piggybacking off of each other's work, but if they're not sharing that work, that can't be happening and I can tell you for sure that nobody is publishing their work unless they think there's an advantage to doing so like getting community contributions like Red Hat. If you mean they saw their competition doing something and decided to compete with them by making their own similar product that does about the same thing, that's not theft and it's probably not meaningful* infringement, unless there's industrial espionage involved, because any two teams making software complex enough to be interesting (or patent-worthy if you prefer) that does the same thing are probably going to end up doing most of the interesting stuff differently. I don't think it's going to end up being meaningful infringement often enough to justify the extra societal costs of legal protection** because it's too hard to copy the internal structure of a software black box without doing something that's already illegal or actionable outside of the patent system. I guess I don't have a problem with processes that involve tangible physical objects because that part of the system seems to work OK. Patenting organisms makes me seriously uneasy - especially since they tend to edit their own genes and do all sorts of fun things like live and possibly reproduce without permission from their 'owners'. *what I think a reasonable developer would consider meaningful - I get that this is a serious weak point in my argument because I have no idea how to quantify it
|
|
#
?
Dec 17, 2013 19:49
|
|
|
Kalman posted:Just off the top of my head: I'm not sure which suits your talking about (these companies have been suing each other for years) or what criteria you use for 'success'. The examiners are (or appear to be) so bad at their jobs that I can't trust that many of these patents should exist, either, but it's my understanding that the court assumes the examiners are competent so it's very hard to make an obviousness or prior art claim, correct? quote:Broadcom v emulex (10gigE) I don't think you can do that without some A/D hardware interface unless you're using an analog computer (but in that case I have no idea what's going on). Also, I'm not sure where you're planning on getting a reliable clock (master signal) without hardware. Otherwise, I think the whole thing could be done in software, but then all you've got is an analog to digital converter hooked up to a general purpose computer that, yes, happens to be doing part of what is described in that patent, but I don't think Broadcom should be able to claim infringement because the computer is just running an algorithm on some data at that point. quote:(Also, a settlement is almost never an admission of validity and I've never seen a lawyer treat it like one. It's a "oh they settled out for nuisance value? Well, let's do our diligence and see what the troll offers us as nuisance value and if it's under the right threshold we will pay it off.") You're thinking like a lawyer here. Most people see settlement and assume guilt. Of course, some companies are going to ask a lawyer who thinks like you, but plenty are just going to pay the nuisance fee (or go out of business) because everyone else gave in, so why should they have any hope.
|
|
#
?
Dec 17, 2013 21:01
|
|
|
Of course, the printer/scanner patents that people have complained about also require hardware (to scan things), the PersonalAudio patents require ancillary network hardware to receive data, a mass storage device to store the data, and an analog audio device to reproduce the audio, etc. Your analysis would exclude patents everyone has problems with from your own definition. In other words - you missed the point again. The point is that there is no good way to define a software patent that does not either exclude things we want to be patentable or else become a dead letter when patents write around it. The problem with bad software patents isn't that they're software, it's that they are bad, and the appropriate (and most likely to be effective) solution isn't to ban software patents but to provide avenues for dealing with bad patents. Most of my troll cases don't involve what you would consider software patents. They're still bad patents. (There is a presumption of validity. It's a problem in jury trials - judges mostly just give it lip service and reexamination more or less preempts it in practice. It's nowhere near the problem you seem to think it is.)
|
|
#
?
Dec 17, 2013 21:33
|
|
|
Kalman posted:(Also, a settlement is almost never an admission of validity and I've never seen a lawyer treat it like one. It's a "oh they settled out for nuisance value? Well, let's do our diligence and see what the troll offers us as nuisance value and if it's under the right threshold we will pay it off.") A big business is going to call their lawyers and do that, but a small business (which is most of them) is going to freak out, settle the suit and stop competing. It happens all the time, and it's probably the most significant problem with the American patent/copyright/trademark system because it creates a huge drag on the economy (patent trolls alone cost Americans $29 billion last year). It's just the cost of doing business for the big companies, but it is a significant competitive barrier to the little ones. You need a corporate patron to survive in that kind of hostile legal environment. Kalman posted:The appropriate (and most likely to be effective) solution isn't to ban software patents but to provide avenues for dealing with bad patents. I've enjoyed reading your comments about the difficulties of implementing patent reform, and the legal mechanics of the cases in question. But I'd be interested in hearing some of your thoughts about how reform should be attempted, and how effective you think that reform would ultimately be in the short- and long-term. For example, what are your opinions about HR 3309 (The Innovation Act): http://www.washingtonpost.com/blogs/the-switch/wp/2013/12/05/the-house-votes-on-patent-reform-today-heres-what-you-need-to-know/ Kaal fucked around with this message at 22:16 on Dec 17, 2013 |
|
#
?
Dec 17, 2013 21:40
|
|
|
I've actually been meaning to write up a summary of the current state of play in patent reform legislation, but haven't had a chance to do so yet due to work commitments. In actual content for the post, there was a Senate Judiciary hearing on patent reform this morning, primarily focused on assertion and litigation issues, that was pretty decent. I mean, it was a congressional hearing so it is going to be what it is, but they had a decent panel (it honestly could have used a rep from a troll to balance things a bit but if I was a troll, I wouldn't step into the line if fits either.). Webcast is archived on the website - http://judiciary.senate.gov
|
|
#
?
Dec 17, 2013 23:37
|
|
|
As a preface: my comments are from the viewpoint of someone who spends probably 65% of their working time defending large companies from trolls, 10-15% of their time prosecuting (obtaining) patents for a range of clients, and the remainder of their time doing a mix of offensive and defensive work in large company v. large company lawsuits. I mostly do high tech work (i.e., cell phones, wifi, and semiconductors) but have enough exposure to the pharma side of things to have a feel for how these bills affect them. My comments are very much from that viewpoint. I've tried to keep the summary of content (normal text) and my slanted viewpoint (italics) separate, but, well, I ultimately am a biased participant in this system. With that, let's summarize some patent reform legislation. I'm going to do a bill at a time because this actually took me about an hour and a half to write up. There's a decent amount in these bills. There's also some miscellaneous technical corrections that mostly fix typos. I'm not going to discuss those. Hopefully you all find these to be helpful; if there's any element of these bills (or of patent litigation in general) that I elided or if there's somewhere where I assumed knowledge, I think if nothing else I've shown in this thread that I am happy to talk at length about patent litigation - ask and I'll try to answer. H.R. 3309 - The "Innovation Act" - the House Bill Author: A Goodlatte bill originally, this has seen some changes since his original draft. (Mostly for the good.) You might remember Goodlatte from such hits as SOPA. Feel free to hate him, but his bill is actually pretty good. Plus, major IP bills of any type always come from the same people, since they tend to be Judiciary Committee leadership/ranking members. There's a reason you repeatedly see names like Leahy(-Smith, a.k.a. the America Invents Act), Hatch(-Waxman, which significantly revamped pharmaceutical patents), and Goodlatte. Status: Reported out of committee 33-5 and passed the House by a vote of 325-91. IP is not generally a partisan issue and the nays are on both sides (although more Democrats.) Has been sent to the Senate and referred to Judiciary, where Leahy will sit on it in favor of his own bill. Assuming Leahy's bill (which I will post about later) passes the Senate, they'll reconcile the differences in conference, unless Leahy and Goodlatte wind up in a similar place on their bills. Section: Pleading Requirements Reform If HR 3309 passes, instead of a bare minimum pleading requirement (which currently amounts to "We think you infringe patent X", plaintiffs will be required to include (in their initial complaint) at least an identification of the specific infringed claims of specific patents, the products/processes/whatevers alleged to infringe identified with as much particularity as possible if no model number is available, what's known as a "claim chart" or a mapping of how each element of the patent maps to an element of the alleged infringing product, and some disclosure reforms (ownership of patent, principal business of patent owner, other assertions of the patent, and whether the patent may be subject to any standard-setting/FRAND processes. This doesn't apply in Hatch-Waxman ANDA cases, but then again, infringement is rarely a major aspect of those cases, since filing the ANDA usually comes pretty close to admitting infringement. This is mostly an unmitigated good. In theory, FRCP Rule 11, which governs conduct before courts, would require all of this anyway. Defendants don't like asking for Rule 11 sanctions out of the gate, partially because the standard is pretty high, partially because if you lose, you're starting your case out with the judge thinking you're whiny. Nobody likes a whiny lawyer. In most courts - the rules are different for each district! - the infringement mapping comes out relatively early in a case, but the standard the bill sets is higher than at least some pleadings I've seen from opposing counsel. The best part: if they don't give you a good pleading, you can just move to dismiss as an insufficient pleading and until the court rules on your 12(b)(6) motion, opposing counsel gets no discovery. This is good because discovery is extremely expensive. Part of the reason discovery is expensive is because general practice firms started doing patent litigation, which means they brought corporate/complex civil discovery practices - discover early, discover often, discover everything - over to the patent world, which had historically just sought discovery on products and inventors for the early phases of the case. All in all, these make the basic bar for filing a case higher. The one downside is that for patents that are really about internal behavior, it becomes harder to make the initial burden, since you won't be able to get technical discovery to show how they infringe. Up until now, at least if you picked your court carefully, you could usually at least get a schematic or source code to review to make your case. That becomes basically impossible to do under these rules. (To which I say: welp.) Section: Form 18 this section is later in the bill, but it's really part of the pleading reforms Form 18 is the form for pleading in patent litigation. It doesn't require very much. This says that Form 18 is gone and we should come up with a new form that includes the above-mentioned information. This is just part of the pleading reforms we already discussed. Nothing new. But seriously, gently caress Form 18 - if it didn't exist, we'd have had heightened pleading already after Twombly/Iqbal. Section: Fee Shifting Basically every one of these patent reform bills has a fee shifting provision, but they're all different in various ways. HR 3309 says that the prevailing party gets reasonable fees and expenses unless their conduct was reasonably justified in law and fact. This will cover attorney's and discovery fees in particular. There are also provisions for requesting certification of ability to pay an award and joining a party and forcing them to pay if the loser can't. One key note: you can join parties with a financial interest, including an interest in the outcome of the case (although there is an exception for attorney receiving compensation related to representation - I am unsure without more research if this affects contingency counsel). Also, if you sue someone - or countersue - and then later give them what's called a covenant not to sue (exactly what it sounds like: a binding legal promise not to sue them) then you will generally be treated as having lost the case. Loser pays! This is basic fee shifting reform. There is, in theory, a fee shifting provision right now, but the other side needs to have been pretty egregious to trigger it. This is mostly good, but does have some problems - problems outweighed by the benefits, though. In practice, fee shifting tends to discourage lovely lawsuits and risk averse plaintiffs. It does tend to discourage suits from small entities, since they're potentially on the hook for my bill if they lose. If the fee shifting provision can drag in contingency counsel, this would probably reduce the use of contingency counsel by trolls since they'll be potentially on the hook in case of a loss, although big trolls like IV or MOSAID will be able to certify payment capability and therefore would still use contingency counsel without any problems. All in all, this basically makes it less likely for bad lawsuits to be brought at the cost of making it less likely for small entities to bring lawsuits. Since most patent litigation has historically been (and likely will be in the future, due to other reforms) relatively large companies suing relatively large companies and trolls suing relatively large companies, I am generally okay with a system that takes that into account and favors a system which has the best outcome for most participants, even at the cost of occasionally causing problems for a Robert Kearns. Section: Discovery Limits (a topic only a lawyer would love, but actually an incredibly useful section overall) Discovery is the process of asking the other side to tell you what they know about various things. This can be anything from "produce a list of all your products" to "search through your corporate email database for information about discussions between yourself and X company." It is, in general, a large proportion of the costs of litigation. For example, of the ~62 million dollars that Apple spent on the offensive side of the Apple-Samsung case, approximately 5 million was spent on discovery costs exclusive of attorney time spent on discovery. Decent bet that somewhere between half and two thirds of the 45 million dollar attorney bill was discovery-related. The reform limits the availability of discovery prior to what's called a Markman hearing. At Markman, the court determines what the words in the patent mean (a process called claim construction). For example, again drawing from the Apple-Samsung case, in one patent the word "applet" was construed. Samsung proposed "a small application designed to run within another program" and Apple proposed "an operating system-independent computer program that runs within an application module". The court decided it meant "An application designed to run within an application module." So, after HR 3309, if the court decides that a Markman hearing is going to be needed, you can't get discovery on anything that isn't information needed in a Markman argument. In practice this will probably mean technical operations documents (which tend to be relatively easy to produce) but not financial documents (expensive and companies don't like producing them), licenses (companies REALLY don't like producing them), standards-related documents, etc. The other important thing about Markman is how often Markman rulings trigger patents being dropped or cases settling entirely. As a result, delaying the expensive parts of the discovery process until after Markman means that you have a chance to backload the fees until after what is often the most important phase of a case. Pushing those fees back until you've got the meaning of the patent in hand and a better evaluation of your risk of infringement means that often, cases will go away cheaper - but the cases that don't go away will be more expensive. It also means that the nuisance cost of a case is lower, since you'll be able to potentially get to Markman and get out for cheaper, which will tend to discourage trolls somewhat. One complication, practically speaking, is that often the Markman constructions get revised over the course of the case - I've worked on one litigation where we got a (third) revised construction order a week before trial. They should probably add in the word 'first' to avoid appeals on the basis of changed constructions invalidating prior discovery, but they'll get slapped down and it's not a big flaw. On a personal note, Markman is a lot of fun, and discovery is incredibly annoying and uninteresting for the lawyers (unless you get off on reading other people's emails and word docs looking for evidence.) So, from my perspective, this is a great reform since it means I get to do more of the interesting parts of cases and less of the terrible. But that is me whining, and no one likes a whiny lawyer. Section: FTC Authority Over End User Demand Letters Identifies deceptive demand letters as particularly concerning and says that demand letters should at least include the patent in question, what is being infringed, and how it is being infringed. Demand letters that don't meet that standard are fraudulent and deceptive and an exceptional circumstance. This section is actually called something about the sense of Congress being blah blah blah blah. Ignore all that. The important words are "actions or litigation that stem from these types of purposely evasive demand letters to end users should be considered a fraudulent or deceptive practice". Those words are magic words that translate to "and the FTC can therefore regulate them." Then there's "exceptional circumstance", which is magic lawyer words for "and the other side will get attorney's fees." The FTC is actually a pretty effective regulator and they tend to aggressively use authorities that Congress has clearly given to them. This section should trigger the FTC doing their very best to destroy the companies sending demand letters to coffee shops with wifi routers and similar. I don't even have a 'good but...' caveat to offer for this, and I don't think I've seen anyone, even on the troll side, who tries to defend those demand letters. I'd have to think about it, but there might be an avenue for using this to argue for invalidity based on patent misuse for these patents if this gets enacted. Section: Pre-Suit Notification and Treble Damages Right now, you can try to get a judgment of willful infringement (and therefore get trebled damages for conduct after the notification) for conduct after a pre-suit notification of the patent. This can be as little as "we own patent X and think you infringe." This section attempts to reform that by limiting effective notifications for willfulness purposes to notifications that have particularity as to what claims are infringed, how they're infringed, etc. This is, all in all, a pretty minor reform, but it makes sense in general. Slightly reduces damages availability, but given the heightened pleading requirements, you're going to need to work up this information anyway. Good, but small. Section: Ownership Transparency Adds some requirements to the mandatory disclosures when a suit is filed (owner; assignee; parties with the right to sublicense; entities with financial interests in the patent; ultimate parent entity.) Also requires the company filing the suit and successors in ownership of the patent to update the PTO if any of this changes after the suit is filed. Failure is punished by removing your ability to recover fees, and the other party can recover any costs needed to discover that information. The PTO already has an assignment database which tracks who owns what patents. In theory. Except participation is voluntary. Which leads to trolls engaging in shell corporation games to hide who owns a portfolio until the day before they file the lawsuit. This is another in the small quality of life reforms - mostly, it shines some light into some of the back end business practices that currently don't show up until a patent is asserted and makes it harder for the IVs of the world to hide what they own. It's also probably why they recently posted a database of their patents online. The recovery of costs provision could in theory be significant if it extends to cost of litigation discovery. In practice, trolls will probably just disclose this information. Good, but minor. Section: Customer Stays This section essentially allows a product manufacturer to step in and litigate on behalf of their customers, if the customer and manufacturer agree and if the customer agrees to be bound by the outcome of the manufacturer's. If they do, then customer litigation is put on hold. I mentioned this provision in an earlier post. I think it's a generally good provision that should fix the remainder of the 'small entities being sued for use of technology' problem. Tiny entities are mostly getting hit with abusive demand letters, covered above. A large portion of the rest of small entities being sued are getting sued for use of technology they bought, not for something they developed. Allowing the manufacturer to step in is helpful. Why would the manufacturer do so? Well. Some people think they won't. I think that Xerox makes a ton of money every day. While patent litigation is expensive, so is marketing and growing market share. A 10 million dollar lawsuit is worth it when you can say to your customers 'Hey, if you buy a Canon you might get sued and they won't help you. We will. Buy a Xerox.' All in all, this shifts the costs and risks away from end users to the manufacturers, which is really where it should be anyway. Section: Judicial Conference Rules Reforms This section asks the Judicial Conference to come up with rules reforms to streamline and improve the discovery process, and directs the district courts and the Court of Federal Claims (the courts where discovery actually takes place - Federal Circuit doesn't deal with discovery) to adopt them. I mean, this really depends on what rules changes come in. They'll probably be helpful, but who knows. I give it an incomplete but a hopeful one, like a teacher who really believes you have it in you to finish that paper before the summer ends and might even go back and change your grade. Section: Protection of IP Licenses in Bankruptcy I have no idea what this section does. It involves Title 11. Title 11 scares me. I like to stay in my Title 35 with some occasional dabbling in Titles 17 and 18, if that's alright. Section: Small Business Outreach, Various Studies Couple sections here on outreach to educate small businesses and on studying various aspects of the patent system. I like studies. This all seems fine. It won't change anything but might provide useful feedback for another reform in a decade. Section: AIA Revisions This section corrects for various aspects of the AIA that need fixing. #1: In inter partes review and post grant review, estoppel is limited to issues actually raised. (Right now, you are estopped from later raising any invalidity defenses you "reasonably could have raised" as well.) This is really convenient for defendants and a huge pain for plaintiffs. What this means is that if you challenge someone's patent at the PTO, then you can't raise the same evidence in court later on. That's fine - no two bites at the apple, etc. Right now you also can't raise anything you reasonably could have brought up at the PTO, which makes it kind of an all or nothing bet to go the IPR/PGR route. After HR 3309, anything you didn't bring up to the PTO is fair game in court. This is a very defendant-friendly provision. I love it, because I do almost entirely defensive work, but it does make life harder for all plaintiffs, worthy and unworthy. That said, the pharma world - which tends to think of itself as a plaintiff - has been supportive of this too. #2: The claim construction standard in IPR/PGR proceedings is to be the district court standard, not the PTO standard. This is a little complex. Remember Markman? There's a set of guidelines for how you figure out what those words mean. At the PTO, they currently interpret claim terms to have "the broadest reasonable interpretation." The district court standard is complicated, but basically has to do with looking at the claim language and specification first, then the prosecution history, then extrinsic evidence, and coming up with the appropriate interpretation based on that. Right now, in IPRs/PGRs - the new processes for PTO review of issued patents - the PTO uses its own standard, which puts a defendant in a weird position where they have to argue for a narrow interpretation in the court and a broader interpretation at the PTO. Everyone I know thinks that this was just a boneheaded omission from the AIA because no one thought the PTO would actually use a broadest reasonable standard in IPRs. In practice, makes the outcomes in IPRs a little bit more like the theoretical outcome in court. In theory this slightly favors plaintiffs, but in practice its preferred by everyone to make things more consistent between PTO and D.Ct. #3: Codify double-patenting rules in first-to-file. There is a bunch of law about how you can't patent the same invention twice, unless you disclaim the term of the later filed patent to extend only to the term of the first filed patent and make it so that if they're not owned by the same entity, the later patent is no longer valid. Problem is, it all relates to the old first-to-invent system. This codifies rules for double patenting in a first-to-file system. Mostly just writing down the rules everyone assumes already applies. Section: Clarifying Jurisdiction This is a provision that is intended to overrule Gunn v. Minton. Gunn said that state courts can hear malpractice suits that relate to patent malpractice. Patent lawyers, being creatures of the federal court system, don't like going to state courts. They have depressing courthouses with bathrooms that don't work. Federal courthouses are really nice. So if you're getting sued for malpractice, the least you can ask is to be able to poo poo in a pristine toilet in an attractive bathroom while you're at court, right? Kalman fucked around with this message at 09:04 on Dec 18, 2013 |
|
#
?
Dec 18, 2013 08:51
|
|
|
IANAL blah blah blah. So the title 11 stuff that Kalman skipped over is interesting. And is a good change if I'm remembering it correctly. When a company goes bankrupt a trustee is appointed to oversee the company. Section 365 (the number I only remember because it's the number of days in a year) allows the trustee to decide whether to continue performing or to reject a certain category of contracts: contracts where both the bankrupt company and the other side of the contract have still have duties to one another. A non-ip example is where in 10 days you're going to buy a car of mine and still owe me the money. If the trustee rejects the contract then the other side joins the pool of creditors demanding money. However, if the contract is relating to intellectual property rights (Except trademarks which are weirdly omitted for some reason) the opposing side can actually force the trustee to continue performing the contract. The license is even forced upon whomever the patent is sold to in the bankruptcy. This is because it would be totally douchey to go in to bankruptcy just to force a higher licensing fee. This would be true for other contracts too, but intellectual property public policy blah blah blah. Also, most other products you can get somewhere else. But a license on the patent can't be - in that respect a patent is a monopoly, cause you usually can't get a license on that patent from someone else. So the change in the law is actually a clarification of what happens when there is US ip involved in a foreign bankruptcy. There is a case working their way through the fed courts right now that prompted this (the company name is Qimonda I think if you want to look it up). In that case, a German? company wants to terminate the licenses on the US patents. Now, the German company isn't in title 11 because it's going through bankruptcy under Germany's laws which don't have a similar provision (in fact I think this provision is somewhat unique to the US). The courts have ruled against qimonda so far (in that they have to keep performing the licenses at least with respect to the US IP), but congress stepping in here will codify that result.
|
|
#
?
Dec 18, 2013 16:34
|
|
|
Ah, okay, so it's just codifying the general rule that licenses run with the patent in the case of foreign bankruptcies. This generally matches up with other circumstances - for instance, if you sell a patent, the purchaser is bound by seller FRAND commitments. Thanks!
|
|
#
?
Dec 18, 2013 16:48
|
|
|
Haha, I just skimmed that section of the bill and the bill also adds trademarks to the categories of IP covered. Edit: kalman, the codification is more directly related to the ability of foreign trustees to reject executory contracts related to IP... But has the effect you mentioned. Meatbag Esq. fucked around with this message at 18:12 on Dec 18, 2013 |
|
#
?
Dec 18, 2013 17:22
|
|
|
So I've been spending the past two days trying to wrestle with rejecting claims in an application that basically says "I want to do X" and the claims say "Do X with a computer" where I can't find art that teaches that anyone would have wanted to do X. I have a serious problem with software patents where stating the problem is its own solution... God, the caselaw sucks.
|
|
#
?
Dec 19, 2013 06:08
|
|
|
Kalman posted:Of course, the printer/scanner patents that people have complained about also require hardware (to scan things), the PersonalAudio patents require ancillary network hardware to receive data, a mass storage device to store the data, and an analog audio device to reproduce the audio, etc. Your analysis would exclude patents everyone has problems with from your own definition. OK - I'm sure there are different analyses that could exclude those, too? I don't expect to solve a problem like this in two sentences. quote:In other words - you missed the point again. The point is that there is no good way to define a software patent that does not either exclude things we want to be patentable or else become a dead letter when patents write around it. The problem with bad software patents isn't that they're software, it's that they are bad, and the appropriate (and most likely to be effective) solution isn't to ban software patents but to provide avenues for dealing with bad patents. So attack the problem from the other end and let them be patented, but craft a law that makes anything done in software exempt from consideration when determining infringement? They get their warm fuzzy IP blanket, but it becomes nigh useless to sue (for instance) a software startup because it's incredibly unlikely that they're infringing outside of the software safe space. It even makes the existing crap harder to use to use! (I have not yet read your analysis of the new law, so maybe that's in there, but I kind of doubt it)
|
|
#
?
Dec 19, 2013 06:12
|
|
|
WhiskeyJuvenile posted:So I've been spending the past two days trying to wrestle with rejecting claims in an application that basically says "I want to do X" and the claims say "Do X with a computer" where I can't find art that teaches that anyone would have wanted to do X. Related, holy poo poo is CyberSource Corp. v. Retail Decisions, Inc. good on 101.
|
|
#
?
Dec 19, 2013 06:47
|
|
|
Munkeymon posted:(I have not yet read your analysis of the new law, so maybe that's in there, but I kind of doubt it) It's not. The new bill attempts to deter patent trolls by making it more difficult (read: expensive) to file and pursue lawsuits for the plaintiff. Some of the reforms I think are ok, because in general they shift fees that the plaintiff would pay anyway to earlier in the lawsuit (e.g., the pleading requirements and the discovery changes). I'm much more hesitant to support the fee shifting provisions, especially the ones in the house bill. These apply not only to trolls but to Joe "better mousetrap" Inventor, whose invention was stolen by Walmart. Full disclosure now that I am actually posting opinions: I primarily do preparation and prosecution work - mostly in the software space, and am currently employed in house by a fortune 500 company. My personal opinions almost certainly do not mesh with my employer on many issues though that may change over time. Side note: I would honestly be willing to bet money that if the supreme court in Alice said that section 101 did not include software patents, that we would have a law passed within a month expressly making them patent eligible subject matter.
|
|
#
?
Dec 19, 2013 06:49
|
|
|
WhiskeyJuvenile posted:Related, holy poo poo is CyberSource Corp. v. Retail Decisions, Inc. good on 101. Why is that holy poo poo? That seems a pretty good ruling unless I missed the point completely (I read it kind of fast). If you can do the computation in your head, saying do that on a computer isn't patentable.
|
|
#
?
Dec 19, 2013 07:50
|
|
|
KernelSlanders posted:Why is that holy poo poo? That seems a pretty good ruling unless I missed the point completely (I read it kind of fast). If you can do the computation in your head, saying do that on a computer isn't patentable. He's saying it's a very good ruling for an examiner supporting rejecting something as unpatentable subject matter.
|
|
#
?
Dec 19, 2013 15:27
|
|
|
WhiskeyJuvenile posted:Related, holy poo poo is CyberSource Corp. v. Retail Decisions, Inc. good on 101. Of course, Cybersource is tough to reconcile with Ultramercial v. Hulu which came down right about the same time with a different panel of judges. (Protip - if you're a patentee with a questionable method patent going to the Federal Circuit, hope and pray you end up with Rader, Newman, and/or O'Malley on your panel - Lourie used to be reliably pro-patent but apparently he got tired of being overruled by the Supreme Court or something) But fear not! The Federal Circuit decided to have an en banc hearing of CLS Bank vs. Alice Corp to sort this matter out once and for all! Unfortunately, that didn't go so well. Well, that's fine - at least we can kind of see where the Federal Circuit is going with this, in that similar claims are probably not patentable, even if they can't articulate exactly why. Maybe they'll take another look at Ultramercial and clarify their reasoning... Ultramercial 2, electric boogaloo! Still patentable after CLS Bank! So now CLS Bank v. Alice has been granted cert, and we all wait with bated breath to see whether and how the Supreme Court will gently caress things up again. Of course, when you look at the claims of Cybersource vs. the claims of Ultramercial, it's easy to see why one set are favored over the other. Cybersource's claims are short, broad, and functional, while Ultramercial's are long and detailed. However, the problem with this analysis is that "level of detail" is a section 112 issue, and bringing it into the eligibility analysis makes for a very difficult test to apply. To go back to your original point about claims that claim "Solving problem x" with a specification that merely recites "X is a problem, someone should solve that", I would suggest rejecting those claims under section 112 first paragraph for lacking written description of what is claimed and 112 second paragraph for being indefinite such that a person of ordinary skill in the art can't figure out what is being claimed. You yourself are ostensibly a person of ordinary skill by virtue of your technical background - if you can't identify what the Applicant is claiming from the claims, they have a 112 problem, not a 101 problem. Also, I'm surprised we don't see more 112 first paragraph rejections from Examiners for using functional claim language. Take a look at Ex Parte Miyzaki. The really good stuff starts around page 21, with the money quotes on page 27, where the board really tears into the use of functional claim language in non-means-plus-function claims. I know the 112 second paragraph rejections in this case have been reviewed by higher courts, but no one ever seems to mention the 112 first paragraph arguments, even though this opinion is deemed precedential and the board specifically raised the arguments as a new grounds for rejection. A colleague of mine is of the opinion that you could argue that "a general purpose processor programmed to do feature x", where one method of doing "feature x" is described in the specification, might be invalid under 112 first paragraph under the reasoning in Miyazaki, since there are probably thousands of different ways to implement feature x, and your claims would be "encompassing any and all structures or acts for performing a recited function, including those which were not what the applicant had invented, [such] that the disclosure fails to provide a scope of enablement commensurate with the scope of the claim." (Miyazaki, page 27). I'm not sure if that argument would be a winner with respect to computer software, since I'd argue the processor itself is sufficient structure, but I think it's an interesting reading and I'd love to see someone make that argument on an appeal to the Federal Circuit just to see what happens. NJ Deac fucked around with this message at 18:52 on Dec 19, 2013 |
|
#
?
Dec 19, 2013 16:29
|
|
|
WhiskeyJuvenile posted:Related, holy poo poo is CyberSource Corp. v. Retail Decisions, Inc. good on 101. CyberSource is good, but only for things where it's "do X with a computer" and X isn't statutory in its own right. You should also look up Fort Props. V. Am. Master Lease as well as Dealertrack v Huber - that's your basic triplet for rejections under 101. Munkeymon posted:OK - I'm sure there are different analyses that could exclude those, too? I don't expect to solve a problem like this in two sentences. Exempting anything done in software from determining infringement is actually even worse than trying to ban software patents. Oh, there's a patent on the process for making cancer drug X? I'll do one of the processing steps in software, now I don't infringe. The point, again, is that software isn't special and shouldn't be treated as special. The problems with software patents aren't unique to software patents - as mentioned, of my current troll cases, zero involve software patents. The solutions aren't unique either (my preferred solution, enhancing when 112-6 applies, would be useful in a lot of bad non-software patents as well). Bad patents are a problem, but they aren't bad because they're software, they're bad because lovely examiners let bad patents through (no offense to Whiskey - any examiner who searches out CyberSource to make a 101 rejection isn't who I am talking about). Trying to eliminate bad patents by banning software patents is like trying to eliminate bugs from your UI code by deciding not to have a UI. Yeah, it will reduce the number of bugs overall but it doesn't actually fix the reason you have the bugs. Kalman fucked around with this message at 16:52 on Dec 19, 2013 |
|
#
?
Dec 19, 2013 16:39
|
|
|
NJ Deac posted:A colleague of mine is of the opinion that you could argue that "a general purpose processor programmed to do feature x", where one method of doing "feature x" is described in the claims, might be invalid under 112 first paragraph under the reasoning in Miyazaki, since there are probably thousands of different ways to implement feature x, and your claims would be "encompassing any and all structures or acts for performing a recited function, including those which were not what the applicant had invented, [such] that the disclosure fails to provide a scope of enablement commensurate with the scope of the claim." (Miyazaki, page 27). I'm not sure if that argument would be a winner with respect to computer software, since I'd argue the processor itself is sufficient structure, but I think it's an interesting reading and I'd love to see someone make that argument on an appeal to the Federal Circuit just to see what happens. I'm trying to get "a general purpose processor programmed to do feature x" interpreted as a 112(f) limitation limited solely to the particular algorithm, and if one isn't disclosed, 112(a) written disclosure.
|
|
#
?
Dec 19, 2013 17:05
|
|
|
also joke's on you, examiners aren't supposed to use caselaw  e: basically, if it's not in the MPEP, we're not supposed to use it, except for whatever's in the 101 guidance, which is terrible, because it a) doesn't give any guidance on how to identify an abstract idea in the first place (which CyberSource is perfect on: "methods which can be performed mentally, or which are the equivalent of human mental work, are unpatentable abstract ideas"), and b) doesn't give any guidance as to how to identify whether the claimed invention preempts every practical application of the abstract idea (  )
WhiskeyJuvenile fucked around with this message at 17:14 on Dec 19, 2013 |
|
#
?
Dec 19, 2013 17:06
|
|
|
I'd love to post this application I'm working on, except a) we're totally not supposed to crowdsource rejections, and b) it hasn't published
|
|
#
?
Dec 19, 2013 17:15
|
|
|
WhiskeyJuvenile posted:also joke's on you, examiners aren't supposed to use caselaw Wait, what? How can that possibly be a good idea?
|
|
#
?
Dec 19, 2013 18:00
|
|
|
KernelSlanders posted:Wait, what? How can that possibly be a good idea? We're not smart enough? MPEP or bust.
|
|
#
?
Dec 19, 2013 18:04
|
|
|
WhiskeyJuvenile posted:also joke's on you, examiners aren't supposed to use caselaw Well yeah and examiners also ignore case law cited in responses.
|
|
#
?
Dec 19, 2013 18:06
|
|
|
gret posted:Well yeah and examiners also ignore case law cited in responses. Yeah, no kidding. I could write all about bad examiners too, just like bad attorneys, bad applicants, bad whatever, but man, I'm just trying to do my job...
|
|
#
?
Dec 19, 2013 18:09
|
|
|
WhiskeyJuvenile posted:We're not smart enough? MPEP or bust. I think it's more that examiners mostly are not lawyers and applying case law to a factual situation is practice of law, and the PTO doesn't want to ask examiners to risk unauthorized practice of law. So instead they have counsel tell examiners what the case law means, which is how we get the MPEP. I would say that if an examiner made a structure/algorithm rejection like the one you made, I would have a hard time arguing against it. Especially as I have made that exact argument in litigation.
|
|
#
?
Dec 19, 2013 18:55
|
|
|
Kalman posted:I think it's more that examiners mostly are not lawyers and applying case law to a factual situation is practice of law, and the PTO doesn't want to ask examiners to risk unauthorized practice of law. Jesus, just give us a reg. number then (I already have one from before I started)
|
|
#
?
Dec 19, 2013 19:23
|
|
|
Kalman posted:I think it's more that examiners mostly are not lawyers and applying case law to a factual situation is practice of law, and the PTO doesn't want to ask examiners to risk unauthorized practice of law. We have a statute, written by lawyers, that you need to be a lawyer to interpret. So, we then get caselaw, written by lawyers interpreting the statutes, that you need to be a lawyer to interpret. So then we get the MPEP, written by lawyers interpreting the caselaw, which I guess you don't need to be a lawyer to interpret, but you sure need to know a lot about patent law. Somehow this is all supposed to make the system function better. I really don't understand the argument that applying the law to a factual situation is practice of law per se. I think there are lots of situations, especially involving government employees, where that's precisely what they do. Building inspectors and administrative hearing officers come to mind, or for that matter anyone in the legislature trying to figure out what the statute he or she is voting on means.
|
|
#
?
Dec 19, 2013 20:30
|
|
|
Also you don't need to be a lawyer in the first place, or we wouldn't permit the existence of patent agents.
|
|
#
?
Dec 19, 2013 20:41
|
|
|
It's funny that after bitching that we focus too much on 101 and not enough on art, I just sent out an action that was 15 pages of 101/112 rejections and 4 pages of art rejections. Patent law!
|
|
#
?
Dec 19, 2013 21:07
|
|
|
KernelSlanders posted:We have a statute, written by lawyers, that you need to be a lawyer to interpret. So, we then get caselaw, written by lawyers interpreting the statutes, that you need to be a lawyer to interpret. So then we get the MPEP, written by lawyers interpreting the caselaw, which I guess you don't need to be a lawyer to interpret, but you sure need to know a lot about patent law. Somehow this is all supposed to make the system function better. Hey, I think examiners should be allowed to seek out case law, I was just offering a possible explanation of why they aren't supposed to. After thinking more about it, it's probably not why - maybe an attempt to ensure more consistency of outcome across the examiner corps? And yes, I do think its fine the the MPEP requires you to know a lot about patent law to interpret. It's intended for lawyers and examiners who are working with patents, not the general public. These are people who in theory should know a lot about patent law. (Also, practice of law is a pretty fine line and your hypos aren't considered practice of law - which has more to do with the practice of law being ill defined and kind of a stupid measuring stick than anything else.)
|
|
#
?
Dec 19, 2013 21:24
|
|
|
|
| # ? May 14, 2024 12:33 |
|
|
Kalman posted:Hey, I think examiners should be allowed to seek out case law, I was just offering a possible explanation of why they aren't supposed to. After thinking more about it, it's probably not why - maybe an attempt to ensure more consistency of outcome across the examiner corps? Yeah, that was kind of my point. I would think they would use the same standard as other professions when determining what is practice of law. Looking at a rash and saying it's poison ivy isn't practicing medicine without a license, but charging for that service or purporting to write a prescription for anti-histamines is. Anyway, speaking of 112, do we know yet how the AIA is going to affect the "best mode" requirement in practice? It seems the act has made that requirement effectively toothless. Presumably the enablement requirement is still in force. Does that ever get patents invalidated in practice?
|
|
#
?
Dec 19, 2013 21:46
|
|