|
You may have known that the IP superstructure was an exploitive system of rent extraction on behalf of a parasitic bourgeoisie but did you know: it's also patriarchal as hell and excludes women both epistemologically and empirically? Check it out!
|
#
?
Dec 9, 2013 21:56
|
|
|
|
| # ? Apr 25, 2024 02:41 |
|
|
Kalman posted:That's the meaning of the prohibition on patenting an abstract idea, with the reason that when you allow that patent, you've foreclosed the field entirely as opposed to foreclosing a method of converting a digit from binary to decimal (e.g using log conversion to change bases as opposed to other methods such as dividing by ten, taking remainder, dividing by 100, taking remainder, etc.). I thought that was the principle complaint -- that patents are granted on problems rather than specific solutions to problems. Essentially this is done by patenting any potential solution. How do you determine if a particular actor has permission to access a particular resource? Well, you check if the actor has permission to access the resource.
|
|
#
?
Dec 9, 2013 22:02
|
|
|
DOCTOR ZIMBARDO posted:You may have known that the IP superstructure was an exploitive system of rent extraction on behalf of a parasitic bourgeoisie but did you know: it's also patriarchal as hell and excludes women both epistemologically and empirically? Check it out! "rent extraction on behalf of a parasitic bourgeoisie"? What the hell does that mean?
|
|
#
?
Dec 9, 2013 22:05
|
|
|
DOCTOR ZIMBARDO posted:You may have known that the IP superstructure was an exploitive system of rent extraction on behalf of a parasitic bourgeoisie but did you know: it's also patriarchal as hell and excludes women both epistemologically and empirically? Check it out! This writer shifts between arguments so often - frequently within a single paragraph - it's virtually impossible to figure out what the author is actually arguing on any level more detailed than your summary. I can't even figure out if it's just bad writing masking a legitimate point or even if there are are legitimate points or if it's just a stream of unconnected thoughts on IP and gender. In any case, if you wanted to check out that subject I sure wouldn't start here. evilweasel fucked around with this message at 22:23 on Dec 9, 2013 |
|
#
?
Dec 9, 2013 22:20
|
|
|
Vladimir Putin posted:Repealing Bayh-Dole would make things worse. Someone has to seek IP protection for things they invent when funded by the government. Otherwise, there will be no investment of capital to develop it further. Without Bayh-Dole it would revert back to the federal government who has almost no institutional connection with the individuals who invented it in the first place. It's going to be a living hell for investors to try to interface with the federal government to try to get access to the inventors and license the technologies. It's public domain if it's funded with public moneys. There, solved. You can build on it if you want, but the work done via the grant is loving public.
|
|
#
?
Dec 9, 2013 22:42
|
|
|
Buffer posted:It's public domain if it's funded with public moneys. There, solved. I assume you're also getting rid of SBIR grants with this policy change.
|
|
#
?
Dec 9, 2013 22:59
|
|
|
Buffer posted:It's public domain if it's funded with public moneys. There, solved. The problem is, public domain ideas don't get turned into products. If it's an idea and not a product, it's not going to help patients. The argument therefore goes that enriching the inventors (whether justified or not) even if the initial research was publicly funded serves public purpose by getting good ideas turned into real treatments.
|
|
#
?
Dec 9, 2013 23:01
|
|
|
Buffer posted:It's public domain if it's funded with public moneys. There, solved. People typically publish and file for patent protection simultaneously. I mean think of the alternative. If you just shove everything out into the public domain without protecting it, the government who paid for the research gets cut out of any of the proceeds. At least with Bayh-Dole if a company uses the published result(that is protected by patent) to create a product, the university gets a cut of the profits. And they can use that share to invest in more research or whatever non-profit educational institutions do.
|
|
#
?
Dec 9, 2013 23:09
|
|
|
Vladimir Putin posted:"rent extraction on behalf of a parasitic bourgeoisie"? What the hell does that mean? Property is theft. Intellectual property is... also theft. Cf. Proudhoun, Marx, etc. The article I linked even has plenty of examples of the people IP law steals from - with a special focus on gender. evilweasel posted:This writer shifts between arguments so often - frequently within a single paragraph - it's virtually impossible to figure out what the author is actually arguing on any level more detailed than your summary. I can't even figure out if it's just bad writing masking a legitimate point or even if there are are legitimate points or if it's just a stream of unconnected thoughts on IP and gender.
|
|
#
?
Dec 9, 2013 23:14
|
|
|
quote:The specific causes of discrimination are also complicated, and elucidated only slightly by the case of Marlo Brown, who identified certain health issues in cats at a shelter she ran and asked a local lab for help. The lab isolated a virus and patented a diagnosis technique, on which she was not listed. She sued. The 1994 ruling (Regents of the University of California v. Synbiotics Corp.) concluded that “Brown was not present and did not participate in any way in the events of the ‘simultaneous conception and reduction to practice’ of the FIV virus,” but merely “brought her sick cats, along with her written observations of the cats’ symptoms, to UC Davis with a suspicion that the cats may have a virus similar to the human AIDS virus.” Her suspicions — and her suggestion of the FIV virus — were the exact ones scientists later confirmed. That this translates into “did not participate in any way” in the process of identifying a virus is confusing, but legally sound. It points to the specific actions prized under patent law, and the narrow, and gendered, set of behaviors it values. Uhh as a scientist, I don't think Marlo Brown deserved to be on a patent for "isolating a virus and patenting a diagnosis technique" when it appears she did neither. Did she take blood samples out of her cats, isolate the virus and then grow them up in media and sequence the genomes? Did she conceive of the antibodies or perhaps PCR fragments that could be used as an actual test for FIV? It has nothing to do with Marlo Brown being a woman, and has nothing to do with people being big meanies. It has everything to do with the authors not understanding what "reduction to practice" means. And the authors also seem to have no idea what it means to contribute substantively and intellectually to a scientific idea.
|
|
#
?
Dec 9, 2013 23:24
|
|
|
Her diagnosis of the cats as having HIV-style symptoms, her notes on those symptoms, were not substantive contributions to the pieces by which that disease is diagnosed?? I'm not on the patent bar so I don't know if that was properly decided - but her contribution was both necessary to the patent being established as well as consisting of substantive work towards the result. Maybe, seeing as we are already discussing changing the patent law, we can take these basic feminist critiques into account? After all we are making decisions about which forms of work to privilege with unique economic power and status. Why not broaden that power and status to previously ignored classes of work (or IP)? Here's an even more incendiary question: why not put the janitors on the patent? Give them some profit sharing, some control over the work their company performs. Wouldn't that be crazy e: To dig a little deeper on your comment, shared by the court, that what Marlo Brown did wasn't science because it didn't involve the use of expensive machines or equipment, that's pretty interesting and basically an example of exactly the kind of epistemological devaluation of the contributions of the sort of people who don't have access to those big fancy machines. DOCTOR ZIMBARDO fucked around with this message at 23:51 on Dec 9, 2013 |
|
#
?
Dec 9, 2013 23:41
|
|
|
DOCTOR ZIMBARDO posted:Her diagnosis of the cats as having HIV-style symptoms, her notes on those symptoms, were not substantive contributions to the pieces by which that disease is diagnosed?? I'm not on the patent bar so I don't know if that was properly decided - but her contribution was both necessary to the patent being established as well as consisting of substantive work towards the result. Maybe, seeing as we are already discussing changing the patent law, we can take these basic feminist critiques into account? After all we are making decisions about which forms of work to privilege with unique economic power and status. Why not broaden that power and status to previously ignored classes of work (or IP)? What does any of that have to do with her being a woman? Is there some argument in there that patent law is discriminatory because I can't find it? Since you seem to get the article's point more than some of us, could you explain please? All I saw was some  about women preferring boy bands and "unoriginal" novels that might, someday, perhaps, hypothetically (but not really) no longer be covered by copyright.
|
|
#
?
Dec 9, 2013 23:48
|
|
|
There's a great conference on the gender issues in IP that American University puts on every single year. Start there if you're interested in the issue. Link: http://www.wcl.american.edu/pijip/go/events/ip/gender
|
|
#
?
Dec 10, 2013 00:31
|
|
|
KernelSlanders posted:I thought that was the principle complaint -- that patents are granted on problems rather than specific solutions to problems. Essentially this is done by patenting any potential solution. How do you determine if a particular actor has permission to access a particular resource? Well, you check if the actor has permission to access the resource. Except that that patent doesn't cover that? It covers: "in response to detecting the request, determining whether said action is authorized based on permissions associated with a plurality of routines in a calling hierarchy associated with said principal, wherein said permissions are associated with said plurality of routines based on a first association between protection domains and permissions." I am not going to gently caress around reading the spec to determine what a calling hierarchy is, but it isn't half as broad as you make it out to be.
|
|
#
?
Dec 10, 2013 00:37
|
|
|
Kalman posted:I am not going to gently caress around reading the spec to determine what a calling hierarchy is, but it isn't half as broad as you make it out to be. A calling hierarchy is a series of functions that call each other. So the operating system calls "main()" and then main calls "printf("hello world");", that forms a hierarchy. In other words, it's a fancy way to make you think they're saying something other than "in an algorithm running on a computer", because that describes every single computer program in existence except for programs that are hyper-trivial (moreso than even hello world). Any computer that has an operating system also has a calling hierarchy, plus every computer that has functions/subroutines, so we're talking about something like "every computer since UNIVAC" in that line. The patent clerks got played for a chump. Paul MaudDib fucked around with this message at 01:22 on Dec 10, 2013 |
|
#
?
Dec 10, 2013 00:59
|
|
|
Kalman posted:Except that that patent doesn't cover that? Every claim I see on there is an overly wordy way to describe anything a computer could possibly be doing in response to the request. I really don't see any limitations in there.
|
|
#
?
Dec 10, 2013 01:02
|
|
|
Paul MaudDib posted:A calling hierarchy is a series of functions that call each other. So you call "main()" and then main calls "printf("hello world");", that forms a hierarchy. Except that that isn't true for this patent. Which is why I said you'd have to read the spec. Instead, a calling hierarchy is defined in the spec by "A calling hierarchy indicates the routines (e.g. functions, methods) that have been invoked by or on behalf of a principal (e.g. thread, process) but have not been exited." So the definition you've proposed doesn't apply and we use that one instead. So it seems to be about patenting authorization detection based on the permissions associated with the hierarchy of functions called but not yet executed by the detected authorization request. If those functions have permission, the thread has permission. Something that does permission detection by, for example, calling a new function to ask if the item in question has permission doesn't infringe. It still seems broad to me, since I think it would potentially read against ACL-type models for file systems which certainly predate the filing date, but again, the patent is nowhere near as broad as you're making it out to to be.
|
|
#
?
Dec 10, 2013 01:23
|
|
|
evilweasel posted:Every claim I see on there is an overly wordy way to describe anything a computer could possibly be doing in response to the request. I really don't see any limitations in there. Looking quickly at PAIR, this application went through at least a couple rounds of amendments before it was allowed. Although the Applicants got a couple of claims allowed in the initial office action, they instead elected to pursue broader claims, since it looks like the initially allowed claims were allowed based on the pencil test (if the claim is longer than a standard unsharpened #2 pencil, the Examiner can't be bothered to try to find all the features and will often just allow it). If you have to add something to your claims to get the application allowed, you don't later get to interpret your claim language broadly enough to get back to the original unamended claims, since you gave that up during prosecution. I'm not motivated enough to review the entire prosecution history and specification to puzzle out exactly what the applicants meant by "wherein said permissions are associated with said plurality of routines based on a first association between protection domains and permissions", but it's worth noting that the Examiner rejected their claims until they added it, and the term "protection domain" is used throughout the specification enough that it's no doubt limited in its interpretation. That's not to say that there aren't lovely patents that get through the office due to idiot/lazy/overworked examiners, but it's very difficult to just look at a patent quickly and say "this is a garbage patent", because interpreting these documents often takes a fair amount of both legal and technical understanding (perhaps this is part of the problem!).
|
|
#
?
Dec 10, 2013 01:31
|
|
|
Kalman posted:Except that that isn't true for this patent. Which is why I said you'd have to read the spec. Please stop while you're ahead. You don't know the faintest thing about the thing you're talking about. What you have described is the exact same thing I did. The function tree does describe functions that are active but have not exited yet. You can find a synopsis here: Note that they are not a part of an Integrated Development Environment, they are an essential aspect of how all modern computers work. If you want, I will go into detail, but the fact is that the thing that specification describes is a trivial aspect of how every program works on a computer that is more modern than 1950 or so. In fact it would be exceedingly difficult and inefficient to construct a program that does not satisfy that criterion. e: I like this representation better. 
Paul MaudDib fucked around with this message at 01:46 on Dec 10, 2013 |
|
#
?
Dec 10, 2013 01:33
|
|
|
Paul MaudDib posted:Please stop while you're ahead. You don't know the faintest thing about the thing you're talking about. What you have described is the exact same thing I did. The function tree does describe functions that are active but have not exited yet. Okay, I misunderstood you then - I thought you were referring to any situation in which going forward a hierarchy of functions is called, rather than the hierarchy up from a current function. My error there. The patent still doesn't cover a prospective analysis - it only applies to doing permission analysis by tracing back through the permissions already granted to the hierarchical functions above the calling function. Calling a new function which says "what permissions are associated with function X?" wouldn't infringe. Like I said, it still seems broad given the ACL example, but it's by no means a patent on "check if permissions are available by checking if permissions are available." (Odds are if I went through the file history or read the spec in more detail I could tell you more limits on what the claim means that further narrow it.) E: And yes, the fact that it would take me an hour or two to puzzle through this (and therefore cost you around a thousand dollars if you were a paying client) to figure out the scope of the claims is a problem, but it isn't the problem the patent was offered for. Kalman fucked around with this message at 01:43 on Dec 10, 2013 |
|
#
?
Dec 10, 2013 01:41
|
|
|
NJ Deac posted:That's not to say that there aren't lovely patents that get through the office due to idiot/lazy/overworked examiners, but it's very difficult to just look at a patent quickly and say "this is a garbage patent", because interpreting these documents often takes a fair amount of both legal and technical understanding (perhaps this is part of the problem!). Oh sure, but I read the patent - and the part Kalman quoted - and I'm not seeing it. Kalman seems to think the restrictions are self-evident. I don't agree, and from looking at the description of code I see something similar to what Paul MaudDib sees: something that looks complex and limited but is describing utterly ordinary mechanisms of how any program would work. I could easily be wrong but I'd like to know precisely where those limits are. The other problem, of course, is that I'm both a lawyer and fairly well versed in technology and I'm not seeing the limits. That tends to get at another problem with patent law: how in the blazes is a loving jury of twelve random people, lead by a judge who probably doesn't have any experience in the area, going to decide? That makes even a lovely infringement case so dangerous.
|
|
#
?
Dec 10, 2013 01:44
|
|
|
hobbesmaster posted:I assume you're also getting rid of SBIR grants with this policy change. Never been anywhere that qualified for these. Vladimir Putin posted:People typically publish and file for patent protection simultaneously. I mean think of the alternative. If you just shove everything out into the public domain without protecting it, the government who paid for the research gets cut out of any of the proceeds. Ahahahahahahahaha, are you loving serious? Government doesn't fund research to make a profit. Vladimir Putin posted:At least with Bayh-Dole if a company uses the published result(that is protected by patent) to create a product, the university gets a cut of the profits. And they can use that share to invest in more research or whatever non-profit educational institutions do. This basically never happens, just FYI. The most frequent thing that happens is the university allows the researcher to spin off a start-up which gets bought up for peanuts, thus basically getting a fee for having their staff poached. Buffer fucked around with this message at 01:51 on Dec 10, 2013 |
|
#
?
Dec 10, 2013 01:48
|
|
|
Buffer posted:Ahahahahahahahaha, are you loving serious? Government doesn't fund research to make a profit. Post better or don't post, thanks.
|
|
#
?
Dec 10, 2013 01:51
|
|
|
Kalman posted:The patent still doesn't cover a prospective analysis - it only applies to doing permission analysis by tracing back through the permissions already granted to the hierarchical functions above the calling function. Calling a new function which says "what permissions are associated with function X?" wouldn't infringe. Right, and let me elaborate on that too. What it actually means by "first association" is almost certainly "first-class function", wherein pointers to functions are treated as an allowable data type. As a simple example, you could have a list of data which takes a pointer to a function. Then you could use pointers to two different functions to cause different actions on different data types. Or, for example, you could store that function hierarchy as a stack of pointers to the current instruction for that function in memory - in other words, when Program() wants to call HelloWorld, it pushes the last address it was at onto the stack, so that when the printf() function is done, the computer knows where to go back to. In other words the function tree would look like this inside the printf function: code:code:I think pretty much all systems do a "prospective analysis" here - you just don't complete the call to SystemInternalFunction() if the CPU isn't operating in supervisor mode. quote:In computer terms, supervisor mode is a hardware-mediated flag which can be changed by code running in system-level software. System-level tasks or threads will have this flag set while they are running, whereas user-space applications will not. This flag determines whether it would be possible to execute machine code operations such as modifying registers for various descriptor tables, or performing operations such as disabling interrupts. The idea of having two different modes to operate in comes from “with more control comes more responsibility” — a program in supervisor mode is trusted never to fail, since a failure may cause the whole computer system to crash. Again, this is a trivial aspect of computers that first became common in the 1970s when multitasking OS's became a thing. You would have to go out of your way to avoid that too. Paul MaudDib fucked around with this message at 02:04 on Dec 10, 2013 |
|
#
?
Dec 10, 2013 01:52
|
|
|
Kalman posted:Unfortunately, PTO guidance only helps at the gate keeping side - until it's either strong Fed Cir (or SCOTUS) precedent or written into the statute, it would still be a litigation problem. http://www.gpo.gov/fdsys/pkg/FR-2011-02-09/pdf/2011-2841.pdf quote:3. Computer-Implemented Means-Plus-Function Limitations:
|
|
#
?
Dec 10, 2013 01:56
|
|
|
Just to correct you, the reason it uses "first association" instead of just association is either because later on (in dependent claims) it talks about a different association or because the examiner issued an antecedent basis rejection saying that association could be confused with the association between permissions and routines in a calling hierarchy. It almost certainly has nothing to do with a first order function and would not be interpreted that way in practice. To elaborate on my prospective analysis point: if the system determines if you have permission by looking at the permissions associated with the functions in your calling hierarchy and granting permission if they have the appropriate permission, that would be potentially a problem. If your system determines if you have permission by looking at the permission already associated only with your function, it isn't a problem (no plurality of routines in a calling hierarchy). If it determines if you have permission by calling a new function that says "does function X have permission?", then no problem. Infringement would arise only for the circumstance where you're looking hierarchically up the tree at each parent and checking the permission for that object to determine if permission exists. There's a grey area which would exist if the permission for the ultimate object is maintained as a result of the hierarchy but your permission checker only looks at the ultimate object's permissions. Infringement in that circumstance would depend on the exact amendments, the file history, etc. Kalman fucked around with this message at 02:04 on Dec 10, 2013 |
|
#
?
Dec 10, 2013 01:59
|
|
|
Buffer posted:Never been anywhere that qualified for these. Except for you know the thing in that post you quoted.
|
|
#
?
Dec 10, 2013 02:02
|
|
|
Kalman posted:Just to correct you, the reason it uses "first association" instead of just association is either because later on (in dependent claims) it talks about a different association or because the examiner issued an antecedent basis rejection saying that association could be confused with the association between permissions and routines in a calling hierarchy. So what exactly is the restriction, if it's not "these set of functions should be allowed to call RestrictedFunction(), others should not"? quote:"in response to detecting the request, determining whether said action is authorized based on permissions associated with a plurality of routines in a calling hierarchy associated with said principal, wherein said permissions are associated with said plurality of routines based on a first association between protection domains and permissions." That exists within every operating system, pretty much. Certainly anything based on the x86 architecture or a "ring" architecture in general, so pretty much everything since the 70s. Certain functions are inside the domain (have supervisor mode enabled) and certain functions are outside the domain (have supervisor mode disabled) and the domain denotes permission to do something. Or again, the private function access domain (only functions inside a class should call), or users who have authenticated themselves (which is the most trivial interpretation, and I believe the one that actually applied here). The permissions don't have to be based on the routines, only associated with them. Paul MaudDib fucked around with this message at 02:23 on Dec 10, 2013 |
|
#
?
Dec 10, 2013 02:03
|
|
|
Paul MaudDib posted:So what exactly is the restriction, if it's not "these set of functions should be allowed to call RestrictedFunction(), others should not"? My guess, based on the description of the file history up thread, is that the applicant specified in more detail what a protection domain is such that rings don't qualify and that's why it got allowed. The spec at least suggests that rings wouldn't qualify since it says protection domains are associations between the source of code and permission, rather than between the executing thread and permission. Since a currently executing thread might be executing unprivileged but the code ultimately was provided from a privileged source, in theory this would automatically elevate privilege since the code is privileged even though the thread isn't. But like I said, I would want to read the file history to really hazard a guess on protection domains.
|
|
#
?
Dec 10, 2013 02:21
|
|
|
evilweasel posted:Every claim I see on there is an overly wordy way to describe anything a computer could possibly be doing in response to the request. I really don't see any limitations in there. That describes just about any computer implemented patent claim ever. I am so loving glad that I examine nearly entirely structural hardware stuff where a belt is a belt.
|
|
#
?
Dec 10, 2013 02:22
|
|
|
Kalman posted:My guess, based on the description of the file history up thread, is that the applicant specified in more detail what a protection domain is such that rings don't qualify and that's why it got allowed. The spec at least suggests that rings wouldn't qualify since it says protection domains are associations between the source of code and permission, rather than between the executing thread and permission. Since a currently executing thread might be executing unprivileged but the code ultimately was provided from a privileged source, in theory this would automatically elevate privilege since the code is privileged even though the thread isn't. But like I said, I would want to read the file history to really hazard a guess on protection domains. Actually it doesn't say the permissions are based on the function hierarchy, it just just says the permissions are associated with a plurality of routines. You don't want the user to have to reauthenticate every time the server makes a function call - you want one user's permissions to percolate through the function calls until you're done handling the whole request. In other words, you need those permissions to associate to a plurality of routines. quote:"in response to detecting the request, determining whether said action is authorized based on permissions associated with a plurality of routines in a calling hierarchy associated with said principal, wherein said permissions are associated with said plurality of routines based on a first association between protection domains and permissions." It's still not really a novel method of determining if someone's authenticated, it's basically "determine if permissions exist (on a computer)". Which is done all the time on every level from basic kernel calls to web servers. Paul MaudDib fucked around with this message at 02:35 on Dec 10, 2013 |
|
#
?
Dec 10, 2013 02:27
|
|
|
Except that still isn't what it means? It's pretty clearly directed to looking at permissions associated with each level of the hierarchy, particularly once you read the description in the specification.
|
|
#
?
Dec 10, 2013 02:50
|
|
|
Buffer posted:Never been anywhere that qualified for these. Everything in this post is wrong. Your concept of what happens in government sponsored research is totally off base.
|
|
#
?
Dec 10, 2013 03:21
|
|
|
Vladimir Putin posted:Everything in this post is wrong. Your concept of what happens in government sponsored research is totally off base. Not to mention that what constitutes "profit" is completely different for a government than it is for a private business. Whenever I hear about government profit I want to stab someone in the mouth. I'm going to stop though so as to not further derail this derail.
|
|
#
?
Dec 10, 2013 03:40
|
|
|
Paul MaudDib posted:Actually it doesn't say the permissions are based on the function hierarchy, it just just says the permissions are associated with a plurality of routines. You don't want the user to have to reauthenticate every time the server makes a function call - you want one user's permissions to percolate through the function calls until you're done handling the whole request. In other words, you need those permissions to associate to a plurality of routines. This isn't about the user per se. This is about the principal associated with certain regions of the call stack based on the origin of the code being executed. So if a program uses code written separately by Alice Inc., Bob LLC and Clara GmbH plus some random crap scraped from Stack Overflow for good measure, each set of code would be part of its own protection domain. Authentication of the principal would be performed when first loading the code by checking the signature of the code. (The related patent filings are about that I believe) Once each principal has been authenticated, the code runs normally. Note that these protection domains can be supplemented at runtime with specific protection domains for users or similar. At runtime when a security check is requested, the stack is scanned from the current frame downward to determine what protection domains exist. I was thrown off by the plurality part since in the implementation the action is forbidden if any principal along the stack lacks the requested permission. This is to prevent Alice's code from calling Bob's code and using it to perform something that Alice's code would otherwise be unable to perform. Bob's code could execute as privileged (the flag in claim 8) in order to execute using its permissions without considering the permissions of the calling code. In that case Alice's code would be able to call Bob's code and it would work so long as Bob has the right permissions. No matter whether Bob is in a privileged context or not, any calls made to Clara's code will require that Bob have the permissions as well as Clara. If Alice were to call Bob's privileged method with a callback (so it's Alice's code again) that's executed within the privileged context, then Alice's code is above Bob's on the stack so any attempt to run using Bob's permissions fails because the security check will hit that before the privileged context. quote:It's still not really a novel method of determining if someone's authenticated, it's basically "determine if permissions exist (on a computer)". Which is done all the time on every level from basic kernel calls to web servers. If you remove all the context that makes something novel then it's not novel any more. No poo poo.
|
|
#
?
Dec 10, 2013 10:28
|
|
|
1337JiveTurkey posted:This isn't about the user per se. This is about the principal associated with certain regions of the call stack based on the origin of the code being executed. So if a program uses code written separately by Alice Inc., Bob LLC and Clara GmbH plus some random crap scraped from Stack Overflow for good measure, each set of code would be part of its own protection domain. Authentication of the principal would be performed when first loading the code by checking the signature of the code. (The related patent filings are about that I believe) Once each principal has been authenticated, the code runs normally. Note that these protection domains can be supplemented at runtime with specific protection domains for users or similar. And, as so often happens in these discussions, an initial discussion of the merits of section 101 patentability of software has turned into a discussion of whether or not the claims of a particular application are obvious and enabled. The problem isn't with whether or not a particular invention is directed to an abstract idea, it's whether or not the invention would have been obvious based on what's known in the prior art. The answer isn't to have special rules for software inventions (because, as previously noted, if you draw a line between software and other inventions, every patent attorney is going to figure out a way to draft their claims so they aren't "software"), but rather we need better examination and more strict disclosure requirements. These claims should be getting rejected based on section 103 and 112, rather than section 101. The main problem is that examiners in the US don't get enough time to conduct a thorough search or to fully understand the applications they are reviewing. The examiners we have are overworked and undereducated. In the EU for example, examiners are expected to have a Ph. D in their respective field, while here it's common for examiners to be hired straight out of college with a B.S. Sure, after a few years in the same art unit they'll probably develop a decent understanding of their assigned technological area, but in the mean time the examination quality suffers. Even then, they aren't allotted enough time to really understand the application in front of them. Most of the objections I hear with regard to software patents are really arguments based on enablement or obviousness. Software isn't so special and magical that it needs special rules - it just needs a thorough enough search and examination so the obvious stuff actually gets weeded out properly at the patent office. Unfortunately we have a Supreme Court that is dead set on conflating obviousness into the subject matter eligibility analysis because there are some claims they just don't like but don't have the right art on record to reject them. This has resulted in a confusing, ambiguous, and unworkable standard that the Federal Circuit has been completely unable to apply consistently. I suspect we'll just see more of the same with CLS Bank, with the Court issuing some half-assed opinion saying "These claims are invalid because their main idea is x, and the prior art doesn't add enough beyond that" without providing any procedure for determining what "x" is, when really the problem is that the Court thinks the claims seem a little too broad but there's no good prior art on record to reject them as obvious.
|
|
#
?
Dec 10, 2013 17:48
|
|
|
NJ Deac posted:And, as so often happens in these discussions, an initial discussion of the merits of section 101 patentability of software has turned into a discussion of whether or not the claims of a particular application are obvious and enabled. The problem isn't with whether or not a particular invention is directed to an abstract idea, it's whether or not the invention would have been obvious based on what's known in the prior art. The answer isn't to have special rules for software inventions (because, as previously noted, if you draw a line between software and other inventions, every patent attorney is going to figure out a way to draft their claims so they aren't "software"), but rather we need better examination and more strict disclosure requirements. These claims should be getting rejected based on section 103 and 112, rather than section 101. I agree with all of this, but would add that the PTO doesn't seem to punish the practice of deliberately wording applications so as to obfuscate what's covered. This practice is part of the problem with software patents, which as you point out should often be rejected on obviousness grounds. A strengthening of usefulness requirements could be helpful too, particularly by requiring novelty in the usefulness of a combination. Rounded corners on tables are useful because the sharp corner doesn't jab your hip. Rounded corners on a phone are useful because they don't jab your hip. There's no novelty in the application on an existing solution to a new problem. I think part of the objection to software patents in particular is that programmers tend to all solve a given problem the same way. Unsurprisingly they would consider this solution obvious. What happens though is, the first company to realize there's a problem will patent the obvious solution to it, and such a patent then grants domain control over the problem -- not any one particular solution to it, the solution to it. The problem then is, how does one prove the solution was obvious, when nobody had tried to solve that problem before?
|
|
#
?
Dec 10, 2013 18:29
|
|
|
Rounded corners on a phone aren't being patented as a new useful feature; they're being patented as a new ornamental feature. Usefulness isn't a requirement for design patents - they don't fall under 101 requirements (new and useful) but instead 171 requirements (new, original, and ornamental).
|
|
#
?
Dec 10, 2013 20:35
|
|
|
Fine, perhaps not a great example, but the second paragraph was my main point.
|
|
#
?
Dec 10, 2013 21:03
|
|
|
|
| # ? Apr 25, 2024 02:41 |
|
|
KernelSlanders posted:Fine, perhaps not a great example, but the second paragraph was my main point. I'd take issue with that paragraph as well. Programmers certainly seem to come up with plenty of different ways to do things like sorting, searching, cryptography, creating pseudorandom numbers, and all sorts of other things. The fact that widespread code reuse and the practice of standard libraries has lead to a culture of programming that doesn't require much in the way of patentable innovation doesn't actually mean that there's a One True Way to do things in most cases.
|
|
#
?
Dec 10, 2013 21:21
|
|