So, are you trying to ban etrade.com and "flipping houses"? Or is risk taking in general ok, and you just want to impose your peculiar morality about playing cards on others?
Kind of a straw man, eh? A key feature of gambling against the house is that, over the long run, the house will always take its cut. The more you play, the more you lose. Not all risk-taking involving money is like that - you can make tons of money in the stock market, for example, if you do it right.
It seems like poker, played against other RL opponents, would be outside the bounds of this, although there's no guarantee that you won't be stuck at a table with a bunch of other people colluding to take your money (or one person playing several seats).
I also had more available cash to "gamble"
A fairly common event in the lives of addicted gamblers is to gamble cash that you don't have available. Since you'd be bankrolling yourself with credit cards rather than really shady loans, you also don't have the demotivating factor of Vinnie showing up to break your kneecaps when you don't pay up.
Why not just call it 1e400.google.com? Screwy domain names with numbers in them make me think of ads, spam, or malware. I'd be a lot more likely to allow javascript/cookies and not put the site in Adblock or the hosts file if it was clearly a Google domain.
It's good to see that somebody around here actually thinks instead of spewing forth uninformed garbage across the tubes.
The department in question handles the legal documentation which forms a record of the assignment of rights in a patent or trademark to another party. Much as with land deed records or other such documentation, the sanctity of these documents must be preserved when they are recorded. That limits the options available for modifying the documents.
Documentation relating to the prosecution of applications (which this isn't) can be entered into the file wrapper regardless of what direction it was placed into the sending fax machine. If necessary, an examiner annoyed with a document showing up upside down when they open it at their workstation can (a) rotate the view temporarily or (b) ask the support staff to rotate the image permanently.
Then it sounds like the issue you're complaining about (aside from, perhaps, a general "patents are bad" argument, which is certainly a respectable opinion in its own right) is that standards are often developed in a way that allows the participating companies to receive patents on the related technology. This is a problem with the way that the standards are developed rather than an inherent problem with the patent system.
Patents filed these days receive a 20 year term from the earliest effective filing date. In this case, the application was filed 24 February 2006. That would mean its expiration date (assuming the assignee pays the maintenance fees) would be 24 February 2026.
However, due to delays by the USPTO, there is a patent term adjustment. The USPTO calculated it at 686 days, which would mean that the patent actually expires (approximately) 11 January 2028.
Finally, a recent court decision (Wyeth v. Kappos) concluded that the USPTO was calculating patent term adjustment incorrectly. This means that the patent may be due a (slightly) longer adjustment. Ultimately, I'm not entirely sure what the correct expiration date is at this point.
In any case, the expiration date is definitely not in 2027.
Definitely an interesting continuation of work being done by various groups over the past couple of decades.
But one thing to note is that crossover isn't especially useful in neural network evolution. In early stages of evolution, it's really no better than random large perturbation of large swaths of the genome. In later stages, it can actually decrease the speed of evolution toward high fitness genomes, because at least some of the time (particularly if there are multiple "species" in the population) crossover ends up being a random large perturbation which hinders the search of local fitness space by mutation; the rest of the time (when individuals from the same "species" are crossed) crossover is no better than mutation.
The reason for this is because the parameters of a neural network are not functional. A section of the genome may correspond to a weight between neurons, but that weight doesn't have a specific function. In biological organisms, each gene is transcribed/translated into a protein, and that protein may have a particular function within the cell. If that gene is acquired by a descendant through crossover, the protein could serve the same (or a somewhat modified) role it served in its parent, even if the rest of the descendant's genome was acquired from the other parent. But with artificial neural networks, the parameters were all evolved as parts of a whole, where each individual parameter has no function on its own, but the behavior emerges from having all of those parameters at the same time.
This could potentially be mitigated by the genome encoding scheme one uses, and of course, if the crossover rate is low enough, the ultimate effect would be small.
There's evidence that the USPTO isn't doing its job at examining patent applications. There are a slew of Federal regulatory services that have been suffering for years from inadequate staffing, the USPTO being one. This seems to be manifesting itself in the examiners not having time to spot a lot of obvious or non-original work.
To some degree this is true, but I want to make it clear that most of us do our job as well as we can in the time provided. Unfortunately, as you mention, we don't have the time needed in some cases. It's taken a lot of arm twisting to get management to agree to modify the way our production is measured, and they were kind enough to give us two more hours per case (sort of) in response to the massive issues with examiner attrition.
But all of this goes back to Congress. The USPTO is supposedly a fee-funded organization. Applicants and patent holders pay us fees for practically everything. When you apply, you pay a fee for filing, as well as a search fee. When the application publishes, you pay a fee for that. When the application is allowed, you pay a fee for it to pass to issue. You pay maintenance fees at up to three points during the issued patent's lifetime. And there are other fees: when we send you a decision on whether your claims are allowed or rejected, you have six months (by statute) to respond, but if you respond after the third month, you have to pay us an extension of time fee. If you want to file a petition pertaining to an app or file an appeal of our decision, there are fees for that as well.
The point is, we charge all these fees, and ostensibly, we run the Office by using the fees we collect... except for two things: one, the USPTO does not have authority to set its own fees. And two, the fees we collect go back into the US Treasury, and Congress has to re-appropriate those fees back to us. Throughout the late 90s, when technology was booming and the USPTO was awash in fees, Congress decided to take those fees and spend them on other programs instead of letting the USPTO use those fees to hire more examiners and update its infrastructure. The result is an inventory of over 700,000 pending applications and an average pendency of about three years - and currently, the Office doesn't have enough money to do anything other than rehire old examiners who left the Office for supposedly greener pastures (because the training expense is vastly reduced, compared to fresh new hires), despite the fact that we're shedding examiners at a rate of 500-600 a year even during the economic downturn.
So if you want a solid and sane patent system, two things are required: reform in the "how patents work" side, and proper appropriations from Congress to ensure that the USPTO can actually get all those applications properly examined.
One final note: patent examination is not expected to have perfect results. Since this is Slashdot, I'll offer a lame analogy: A person walks up to you and says, "I may have lost my needle in that giant haystack over there, but I'm not sure. Could you try to find it for me, please?" Not only is searching the prior art like looking for a needle in a haystack, but you aren't sure the needle even exists, and there's a huge line of other people who want you to look for their needles as well. Eventually, you may have to just give up and tell the person that you didn't find their needle (i.e., give them a patent). While occasionally we do issue something that shouldn't be issued, there are thousands of other applications that go abandoned and thousands more in which the claims are narrowed over the course of prosecution to ensure that the issued patent covers only what the applicant actually invented. As with everything else, though, it's the screw-ups that get all the press.
Which doesn't stop patent claims covering "A machine readable medium containing X", where X is some thing not patentable on its own.
Actually, current policy at the USPTO is to reject such claims by not lending any patentable weight to X in cases where X is nonfunctional descriptive material. So, if a claim recites "A compact disc embodied with Rick Astley's 'Never Gonna Give You Up'", the examiner only needs to find prior art describing a CD, because the music doesn't change the functionality of the medium.
I agree (more or less) with your other statements, though.
Actually, the patents can't cover the bitstreams. Case law such as In re Warmerdam indicates that nonfunctional descriptive material (such as a picture or a song) isn't a process, machine, article of manufacture, or composition of matter, so it doesn't meet the requirements of 35 USC 101.
However, the patents can potentially cover (modulo any prior art issues) the process of transmitting the bitstreams. It may sound like some really thin slices of salami to you, but that's how the law works.
By the way, your example about the "fancy new saw" and the resulting cut wood is wrong. As long as the cut wood resulting from using the fancy new saw has a specific, substantial, and credible use (for example, it's not merely ornamental, although there's something called a design patent that covers ornamental designs), it is patent eligible under 35 USC 101, because it's an article of manufacture.
It might account for the first vertical stripe directly (ten days' worth of minimal packet data accumulated into one data point), but then you would expect the data from the busy countries to then be ten times as high for that one data point.
But what it does indicate is that there are ten days of missing data that most likely show the start of this behavior and could provide further insight.
I wonder whether this data was inadvertently left out by the submitter, inexplicably dropped by the third-party processing company, or intentionally deleted from the server logs by some outside party who gained access to the box.
Slashdot Mirror
I thought that shields + lasguns = bad stuff happens.
Mark Hamill is probably just bitter because he never did get those power converters.
That's okay - Mac users got confused for the longest time when presented with a two-button mouse.
Sometimes even with a one-button mouse, such as when the iMac zero-button "HULK SMASH" mice were in vogue.
and besides who wants to argue with ninjas?
Uh........ pirates?
So, are you trying to ban etrade.com and "flipping houses"? Or is risk taking in general ok, and you just want to impose your peculiar morality about playing cards on others?
Kind of a straw man, eh? A key feature of gambling against the house is that, over the long run, the house will always take its cut. The more you play, the more you lose. Not all risk-taking involving money is like that - you can make tons of money in the stock market, for example, if you do it right.
It seems like poker, played against other RL opponents, would be outside the bounds of this, although there's no guarantee that you won't be stuck at a table with a bunch of other people colluding to take your money (or one person playing several seats).
I also had more available cash to "gamble"
A fairly common event in the lives of addicted gamblers is to gamble cash that you don't have available. Since you'd be bankrolling yourself with credit cards rather than really shady loans, you also don't have the demotivating factor of Vinnie showing up to break your kneecaps when you don't pay up.
Why not just call it 1e400.google.com? Screwy domain names with numbers in them make me think of ads, spam, or malware. I'd be a lot more likely to allow javascript/cookies and not put the site in Adblock or the hosts file if it was clearly a Google domain.
He said for deeper work they use a laser.
That sounds awesome, but I'm a bit confused: how do they fit the whole shark inside your mouth?
Yes, but does it shift paradigms outside the box?
What happens to open source browsers like FF who can't pay for the patents and licenses?
Maybe HTML5 in Firefox should mean that I can right click and "save as". Then it won't really matter.
And I mean the US government, including the whole political system, but not all people living there.
What? You probably shouldn't trust us regular folks. We eat babies over here, you know!
I also love how you're supposed to look to the claims to find out what the patent covers instead of just reading the title and stopping there.
It's good to see that somebody around here actually thinks instead of spewing forth uninformed garbage across the tubes.
The department in question handles the legal documentation which forms a record of the assignment of rights in a patent or trademark to another party. Much as with land deed records or other such documentation, the sanctity of these documents must be preserved when they are recorded. That limits the options available for modifying the documents.
Documentation relating to the prosecution of applications (which this isn't) can be entered into the file wrapper regardless of what direction it was placed into the sending fax machine. If necessary, an examiner annoyed with a document showing up upside down when they open it at their workstation can (a) rotate the view temporarily or (b) ask the support staff to rotate the image permanently.
Then it sounds like the issue you're complaining about (aside from, perhaps, a general "patents are bad" argument, which is certainly a respectable opinion in its own right) is that standards are often developed in a way that allows the participating companies to receive patents on the related technology. This is a problem with the way that the standards are developed rather than an inherent problem with the patent system.
This will truly revolutionize the game of Mornington Crescent!
The veto power of patents equates to the right to prohibit me, and everyone, from writing a functional video player.
Yep, that's pretty much what patents are for.
It seems like the Daleks are always doing that. That gods Jon Pertwee was there before. What are we going to do now?
I suggest reversing the polarity. It's always worked before.
Patents filed these days receive a 20 year term from the earliest effective filing date. In this case, the application was filed 24 February 2006. That would mean its expiration date (assuming the assignee pays the maintenance fees) would be 24 February 2026.
However, due to delays by the USPTO, there is a patent term adjustment. The USPTO calculated it at 686 days, which would mean that the patent actually expires (approximately) 11 January 2028.
Finally, a recent court decision (Wyeth v. Kappos) concluded that the USPTO was calculating patent term adjustment incorrectly. This means that the patent may be due a (slightly) longer adjustment. Ultimately, I'm not entirely sure what the correct expiration date is at this point.
In any case, the expiration date is definitely not in 2027.
This is sladshot, what do you expect?
I except prefection!
Definitely an interesting continuation of work being done by various groups over the past couple of decades.
But one thing to note is that crossover isn't especially useful in neural network evolution. In early stages of evolution, it's really no better than random large perturbation of large swaths of the genome. In later stages, it can actually decrease the speed of evolution toward high fitness genomes, because at least some of the time (particularly if there are multiple "species" in the population) crossover ends up being a random large perturbation which hinders the search of local fitness space by mutation; the rest of the time (when individuals from the same "species" are crossed) crossover is no better than mutation.
The reason for this is because the parameters of a neural network are not functional. A section of the genome may correspond to a weight between neurons, but that weight doesn't have a specific function. In biological organisms, each gene is transcribed/translated into a protein, and that protein may have a particular function within the cell. If that gene is acquired by a descendant through crossover, the protein could serve the same (or a somewhat modified) role it served in its parent, even if the rest of the descendant's genome was acquired from the other parent. But with artificial neural networks, the parameters were all evolved as parts of a whole, where each individual parameter has no function on its own, but the behavior emerges from having all of those parameters at the same time.
This could potentially be mitigated by the genome encoding scheme one uses, and of course, if the crossover rate is low enough, the ultimate effect would be small.
Why not create an agency with that job
I'll ask 'eem, but I don' think he'll be very keen... we've already got one, you see!
There's evidence that the USPTO isn't doing its job at examining patent applications. There are a slew of Federal regulatory services that have been suffering for years from inadequate staffing, the USPTO being one. This seems to be manifesting itself in the examiners not having time to spot a lot of obvious or non-original work.
To some degree this is true, but I want to make it clear that most of us do our job as well as we can in the time provided. Unfortunately, as you mention, we don't have the time needed in some cases. It's taken a lot of arm twisting to get management to agree to modify the way our production is measured, and they were kind enough to give us two more hours per case (sort of) in response to the massive issues with examiner attrition.
But all of this goes back to Congress. The USPTO is supposedly a fee-funded organization. Applicants and patent holders pay us fees for practically everything. When you apply, you pay a fee for filing, as well as a search fee. When the application publishes, you pay a fee for that. When the application is allowed, you pay a fee for it to pass to issue. You pay maintenance fees at up to three points during the issued patent's lifetime. And there are other fees: when we send you a decision on whether your claims are allowed or rejected, you have six months (by statute) to respond, but if you respond after the third month, you have to pay us an extension of time fee. If you want to file a petition pertaining to an app or file an appeal of our decision, there are fees for that as well.
The point is, we charge all these fees, and ostensibly, we run the Office by using the fees we collect... except for two things: one, the USPTO does not have authority to set its own fees. And two, the fees we collect go back into the US Treasury, and Congress has to re-appropriate those fees back to us. Throughout the late 90s, when technology was booming and the USPTO was awash in fees, Congress decided to take those fees and spend them on other programs instead of letting the USPTO use those fees to hire more examiners and update its infrastructure. The result is an inventory of over 700,000 pending applications and an average pendency of about three years - and currently, the Office doesn't have enough money to do anything other than rehire old examiners who left the Office for supposedly greener pastures (because the training expense is vastly reduced, compared to fresh new hires), despite the fact that we're shedding examiners at a rate of 500-600 a year even during the economic downturn.
So if you want a solid and sane patent system, two things are required: reform in the "how patents work" side, and proper appropriations from Congress to ensure that the USPTO can actually get all those applications properly examined.
One final note: patent examination is not expected to have perfect results. Since this is Slashdot, I'll offer a lame analogy: A person walks up to you and says, "I may have lost my needle in that giant haystack over there, but I'm not sure. Could you try to find it for me, please?" Not only is searching the prior art like looking for a needle in a haystack, but you aren't sure the needle even exists, and there's a huge line of other people who want you to look for their needles as well. Eventually, you may have to just give up and tell the person that you didn't find their needle (i.e., give them a patent). While occasionally we do issue something that shouldn't be issued, there are thousands of other applications that go abandoned and thousands more in which the claims are narrowed over the course of prosecution to ensure that the issued patent covers only what the applicant actually invented. As with everything else, though, it's the screw-ups that get all the press.
Which doesn't stop patent claims covering "A machine readable medium containing X", where X is some thing not patentable on its own.
Actually, current policy at the USPTO is to reject such claims by not lending any patentable weight to X in cases where X is nonfunctional descriptive material. So, if a claim recites "A compact disc embodied with Rick Astley's 'Never Gonna Give You Up'", the examiner only needs to find prior art describing a CD, because the music doesn't change the functionality of the medium.
I agree (more or less) with your other statements, though.
Actually, the patents can't cover the bitstreams. Case law such as In re Warmerdam indicates that nonfunctional descriptive material (such as a picture or a song) isn't a process, machine, article of manufacture, or composition of matter, so it doesn't meet the requirements of 35 USC 101.
However, the patents can potentially cover (modulo any prior art issues) the process of transmitting the bitstreams. It may sound like some really thin slices of salami to you, but that's how the law works.
By the way, your example about the "fancy new saw" and the resulting cut wood is wrong. As long as the cut wood resulting from using the fancy new saw has a specific, substantial, and credible use (for example, it's not merely ornamental, although there's something called a design patent that covers ornamental designs), it is patent eligible under 35 USC 101, because it's an article of manufacture.
Er... sorry, the x axis goes by hours mostly, so it would be 240 times as high rather than 10.
It might account for the first vertical stripe directly (ten days' worth of minimal packet data accumulated into one data point), but then you would expect the data from the busy countries to then be ten times as high for that one data point.
But what it does indicate is that there are ten days of missing data that most likely show the start of this behavior and could provide further insight.
I wonder whether this data was inadvertently left out by the submitter, inexplicably dropped by the third-party processing company, or intentionally deleted from the server logs by some outside party who gained access to the box.