it is MY copy that is being transferred over the network, I just bought it. I don't need a license to transfer MY copy over the network to MY machine, just as I don't need a license to make the transient copies necessary to listen to a CD I stick in my CD player.
I may have to agree to certain terms in order for Apple to agree to sell me music through iTMS, and Apple may be required by THEIR license to require me to agree. To the extent that I want to do something that copyright doesn't allow the owner of an authorized copy, I'd need a license (sub-licensed from Apple), but I don't need a license just to receive the song from Apple and play it. Burning a playlist to a CD and being allowed to give that to someone else is one such thing I'd need a license to do, for example.
But it's not insane to say that I am remotely controlling Apple's servers? I didn't direct my machine to do anything, I clicked on a button that said Pay The Man, then The Man sends me a song that shows up on my disk. Without Apple's actions the copy wouldn't appear. We are both doing it, and Apple has a license to send me a copy. If neither of us had a license, then both would be responsible for the infringement, but that's not the case, there is no infringement because it is an authorized copy.
Yes, the Napster case says that the user downloading a copy is infringing by making a copy. That's because the person uploading the file wasn't authorized to distribute it. If I buy a copy from an authorized distributor, however, who is doing the copying is irrelevant, because it is NOT INFRINGING.
You do have a point that such a tiered service would give an incentive to the carriers to provide poor service so that customers will upgrade - however, I don't see a problem with this kind of "tiering" in general. If I want to pay more to get better service, that's fine. It is the discrimination based on destination or content or application that is the problem.
I have no problem with setting things up so that connecting to a server that pays more to THEIR ISP for better service gives me a faster connection TO THEM, nor do I have a problem with having it faster if I pay more to MY ISP for better service (to everyone, or even to a specific service). The problem is when my ISP wants to charge Google a fee in order to allow ME to connect to Google faster, when Google isn't a customer of theirs, by intentionally degrading the connection if they don't.
Some forms of discrimination are fine: if AOL or Verizon or whoever wants to make a special deal with Google to set up private lines and caching servers to both reduce their costs and improve their customer's connections to Google, and they want Google to share in the cost of that (and presumably it would reduce Google's costs to provide the same level of service), that's fine. That might be a reason to choose one ISP over another. That's a partnership. Distinguishing that from intentionally degrading connections is the difficult part.
Exactly. It doesn't cost more to transfer bits containing video, VOIP, e-mail or web pages. You might need a different QoS for something like VOIP, so charge for special treatment (but don't allow intentionally degrading service just so you CAN charge to not mess it up). Charge by peak bandwidth, delay, jitter, since that's what costs you money - but each individual bit doesn't cost anything. Put in intelligent, fair throttling so everyone gets their share of available bandwidth and you don't need caps. Since it is statistical behavior with a large number of independent actors, it is easy to assign a cost without measuring each individual bit, tracking where it is going, or charging based on what kind of bit it is.
It is only when the bit carriers get into the value-added bits as well that you get a conflict of interest and no longer have an incentive to create an efficient network. That they think they can get away with it demonstrates that they DO have monopoly power, and shouldn't be allowed to do it. If it was truly free market, there would be no problem with them offering a network connection subsidized by a VOIP or music or video service, but the "last mile" issue pretty much forces it to NOT be a free market as long as the carrier is also allowed to offer services other than transport.
I don't misunderstand what a copy is at all. The courts have held that a temporary copy in RAM is "a copy". I am asking Apple to, through their server and their software, to make a copy of a song on my disk. Since they ARE AUTHORIZED to do so, what's the problem?
A band writes a song, performs it, records it, and allows me to order a copy in one of two ways: I can download it or I can have them burn a copy to disk and mail it to me. How is there a difference? When I request that they burn a copy to disk, am I making the copy (since it is my request that causes them to burn it)? The argument is ridiculous. Granted, if it is an UNAUTHORIZED copy, then I don't have any rights to it at all - since I never owned an authorized copy, I can't make any of the the intermediate copies necessary to create a permanent copy on my hard drive. With a song I purchase from iTMS, Apple makes the initial AUTHORIZED copy (which I now own), so there shouldn't be any problem with making intermediate copies to transfer it to my machine, regardless of who initiates it. I don't need a license, Apple does.
Actually, the courts were ruling that you did NOT have a right to make the temporary copies in RAM needed to run a program, which is why copyright law was changed to grant the owner of an authorized copy the right to copy and modify the software to the extent necessary to make it run. This would include copying it from CD to hard drive to RAM, for instance.
If the program won't run unless I click I Agree, and I Do Not Agree with the terms, then patching the program so it works without clicking I Agree might be a "necessary step" to get it to run!
If I reject the license, even if "it said so on the box", I am still the owner of that copy, and can use the content in ways that copyright law allows me.
A terms-of-use agreement on a server is a different matter. You clearly have no right to download content or connect to a server without permission, so if it says that in order to continue you have to agree to certain terms, then those terms ARE enforceable. It is a completely different situation from having an authorized copy of some software in your possession that you own.
The point is, it is an AUTHORIZED COPY. Apple has a license to authorize me to make that copy, if in fact that's what is happening. I don't need a persistent license to play it once I have it, all I need is permission from someone authorized to give it, which is Apple.
I don't understand how you can have it both ways. If Napster was "illegally distributing", then surely a "legal distributor" such as iTMS isn't causing the recipient to make an unauthorized copy!
Look at it from a different standpoint - a reasonable model for selling music in the physical world would be for the store to burn a CD on demand, rather than keeping inventory. If they are authorized to make such copies, and I go in to the store to request they make such a copy, it isn't ME that is making the copy, it's the store. They're the ones authorized to do it, so I don't need a license in order to be allowed to ask them "please, sir, may I have some more?" In the Napster case, SOMEONE was guilty of infringing by making unauthorized copies, and the court rejected the argument that "no, I'm not making copies, he's making copies" by both sides, plus found the man-in-the-middle Napster liable as well. They were all participating in the making of an unauthorized infringing copy. In the case of an authorized copy, you don't need to figure out who actually made the copy, whether it was me by asking Apple's servers to send a copy, or Apple by granting my request, because it is AUTHORIZED.
I modify the license to change terms I don't agree with, initial it, then agree to it. Never yet a license that I couldn't agree to after doing that!
If it is a click-license in the software, patch it to change the terms first. Since you haven't yet agreed to the terms that say you can't modify or reverse engineer it, there's no reason you can't do that. Since most software click-licenses show the terms in a standard scrolling text box, you might even be able to make a standard extension to allow you to modify the terms on the screen before you click "I Accept".
Or just patch it so that the buttons for "I Accept" and "I Don't Want To Run The Software That I Paid Good Money For" are reversed. After all, you don't need a license to run it, unless you want to do things that the license allows that you don't already have the right to do.
Generally, it just says "All rights reserved Copyright xxxx Studio Name Here"; it may say "licensed for home use only", but that license must be referring to the license the owner of the copyright gave to the distributor. It isn't a license to me, and I don't need a license.
Copyright law lists certain limited rights that a copyright owner has. If it isn't an explicitly listed right, then the copyright owner can't control it. Non-public "performance" is NOT a right that the copyright owner has control over; that is a right that the owner of the "authorized copy" has (performance for software being "using it"; for a DVD or CD watching and/or listening to it, etc). You don't need a license to read a book, either, but doing a public reading of any substantial portion of it would be a violation of the copyright owner's rights. That doesn't mean you can't read it to your child!
The "unauthorized copying or distribution of this copyrighted... is illegal..." messages aren't licenses, nor are they licensing restrictions. They are merely helpful reminders to you of what copyright law says. When it talks about "unauthorized lending", etc, it has little force - it is authorized if copyright law allows it, if copyright law doesn't allow it then it was already an infringement to do that, even if it didn't say so. At best, it could be used to demonstrate that no license was given or implied.
Actually, with DVRs reporting (in a suitably anonymous way) how many people watch a particular ad and how many skip it, they could provide a valuable service to their advertisers, giving them real feedback on what type of advertising actually works. They should embrace the technology, not fight it!
Uh, no? I was responding to your statement that they weren't talking about "direct descendants", but "somehow related" - you said specifically:
one person who, NOT NECESSARILY DIRECTLY is related to all people currently alive. Ie he's not your grandfather, he's that weird great uncle on your mother's side
I'm disagreeing - if you were related to that person by him being an uncle to one of your direct ancestors, just go back one generation and you have someone who is a direct ancestor! Makes no difference, another 30 years or so.
The rest wasn't necessarily disagreeing with you, but expanding on it.
You're right that they're not saying there was an "Adam" 5000 years ago, but they ARE saying that there was a most recent direct descendant to everyone alive today. That doesn't mean that person is the ONLY person alive at the time who had descendants, nor that if you go back one or two more generations, there aren't LOTS of common ancestors, just that one of them has to be the most recent. Going the other way, you will eventually come to a time when EVERYONE then alive was either an ancestor of everyone now alive, or has NO living descendants in this time. Ultimately, you will come down to one individual who is THE common ancestor of everyone alive today, but they aren't saying that was 5000 years ago.
It is still "prior art", regardless of who used or disclosed it. The inventor gets a year after such disclosure to patent it. Anyone else would have to prove that they invented it first, and in any case lose the ability to patent it after the year is up, even if they did come up with it first.
Point is, there are many ways to implement intermittent wipers; claim 1 just references basically any timing circuit using a capacitor and resistor, which is an obvious way of implementing a timing function, even in 1969. My point about the rain sensor was not whether it could have been implemented back then, but that the basic "idea" of a wiper that could not go just slow or fast, but also intermittently, regardless of the details of how to implement it, should not be patentable - as it was obvious even to a kid back then.
I just don't see the point of being able to patent something that any engineer would come up with when asked how to do something.
I don't think one should be able to get a patent on the idea of an intermittent wiper. A particular implementation, sure. I know that before I ever saw an intermittent wiper, I had the idea of a wiper that would automatically do one wipe based on a sensor that would determine how fast the rain was falling. Doing it manually is actually a step BACKWARD from what I "invented". It's easy to come up with novel ideas, it's the implementation that can be hard, and that's what should be protected.
Prior art doesn't have to be published to invalidate a patent, it only has to be used "in public". That is, if you invent something (say, a novel engine in a car), and you use the invention in public, you can lose the right to patent it a year later even if you haven't revealed the details. A technique used in a program, or even a server, that can be used by people other than the inventors (e.g. people who haven't signed an NDA) can be used as prior art, even if the code implementing the technique is never made public.
This patent was filed in 1998, with a priority claim from an earlier application from Sep. 1997, so anything in public use, or described, or documented, before Sep. 1996 is relevant. In one of TFAs (I think from the other story about Bruce Perens), someone commented that Postgres95 seems to match up with this patent, along with TopLink in Smalltalk in 1994.
I was thinking more about the case where somebody comes back the next day, sees that the bid is well past their original maximum bid point (with several active bidders) and wonders why. You're right that with only one other bidder in an auction, you can't find out how high they're willing to go (which is how they then end up bidding it up a couple bucks at a time repeatedly, trying to find out, but without committing to a higher price - the need to know how much someone else is willing to pay gets them to ignore their own judgement, plus maybe get some sort of vicarious thrill of getting the other guy to pay more for it; in that game, you actually lose when you win the auction).
If 20.53 wins, then 21.00 would have won as well. Ok, so you might save 47 cents. The real problem is, there is no fixed line where it all of a sudden becomes "too expensive". If 21.00 is your maximum, are you SURE you wouldn't buy it for 21.01? It's only another penny. Come on, sir, just one wafer-thin mint? Which grain of sand is the one that changes it from "some sand" to "a pile of sand"? Which strand of hair do you lose that you all of a sudden become "balding"? At which particular second do you become "old"?
It isn't that the bidding is going on for days, but that the item is available for new people to find it and bid on it. The first minute is exactly the same as the last minute, if you have the highest bid.
Sniping won't hurt the rational bidders who put in an early bid at their true maximum price. The only ones who hurt that bidder are the ones who change their concept of how much they're willing to pay based on how much others are bidding (which isn't ENTIRELY irrational; that someone is willing to pay 20% more than you thought it was worth may be an indication that you undervalued it). Sniping avoids THOSE bidders, not the ones who put in their true maximum bid at the beginning, and since there are many auctions where those will be the only bidder, it makes sense to do it.
It works because people often DON'T bid their maximum; they bid a lower amount, and thus if you snipe (and they don't), you get it for less than the maximum they'd actually be willing to pay. The other case is when they THINK they have bid the maximum, but when they see someone else also wants it and has a higher bid, they re-evaluate how much they're willing to pay for it. By sniping, you avoid triggering their irrationality.
However, if eBay wants to allow for sniping, they should simply allow people to put in sealed bids.
You make a good point, but eBay could easily make this easy to deal with. You could make a bid be conditional on being outbid in another auction, create a whole chain of bids, and automatically order them based on auction closing times. They could even handle and resolve cases where multiple auctions close at exactly the same time, and apply first-to-bid rules there as well (among conditional bids only; already-bid non-conditionals would still be "earlier" than any conditional bids, although a conditional bid might still bump up the price at the very end, and thus would have to show up in the bidding history to explain the ending price).
One thing to note, in the commentary in the article, they mention that the study showed an average of fewer than 2 bids per auction. Sure, you're going to win lots of auctions if you bid on items that no one else was interested in (and thus only make one bid), and are much less likely to win an auction if you bid on items that other people are bidding on (especially if you bid and re-bid multiple times, indicating you aren't really sure exactly how much you're willing to pay for it).
Re:That's partly because of our consumer-oriented
on
How to Win on Ebay: Snipe
·
· Score: 2, Insightful
eBay could fix this by a) having an auction extended by an hour if any bids come in during the last hour; b) not allowing anyone to bid during that last (extended) hour if they hadn't already bid at least once. This would tend to make people bid on an item earlier, which (since people are bidding irrationally) will tend to drive up prices, which will result in increased fees for eBay.
From the commentary in the USA Today article, they do acknowledge that sniping works because many bidders are naive or irrational. By sniping, you avoid giving these irrational bidders information as to how high you'd go, thus preventing them from overbidding what they would rationally bid.
The other good point in support of sniping is when you're trying to get one item, there are several auctions, and you don't want to commit to any one of them. However, I'd think a better strategy (absent the irrational bidders) would be to bid your maximum on the earliest-to-close auction, then bid your maximum in the next one whenever you get outbid on the first one, and so on. Again, eBay could allow you to make conditional/chained bids like this automatically.
Since sniping is basically just a sealed-bid auction, you're still going to get outbid by an idiot who wants to pay too much. The simple fact that I'm willing to pay more than the current next-highest bidder isn't really giving away a whole lot of information. They still have no idea how high I'm willing to go except to commit to paying more than it is currently. If they're willing to commit to paying more than I am willing, they were going to outbid me with a snipe even if I try to snipe it as well. However, as the article makes clear, what really is happening is that someone sees you want it at a certain price, and that actually changes what they're willing to pay for it.
It doesn't surprise me at all that "the more bids you make on an item, the less likely you are to win it", since you shouldn't ever need to make more than one bid, at the very beginning. If you do get into re-bidding, a) you're irrational and b) you're bidding against someone else irrational, and in that case you certainly are better off avoiding their irrationality and concealing your irrationality by sniping.
Slashdot Mirror
it is MY copy that is being transferred over the network, I just bought it. I don't need a license to transfer MY copy over the network to MY machine, just as I don't need a license to make the transient copies necessary to listen to a CD I stick in my CD player.
I may have to agree to certain terms in order for Apple to agree to sell me music through iTMS, and Apple may be required by THEIR license to require me to agree. To the extent that I want to do something that copyright doesn't allow the owner of an authorized copy, I'd need a license (sub-licensed from Apple), but I don't need a license just to receive the song from Apple and play it. Burning a playlist to a CD and being allowed to give that to someone else is one such thing I'd need a license to do, for example.
But it's not insane to say that I am remotely controlling Apple's servers? I didn't direct my machine to do anything, I clicked on a button that said Pay The Man, then The Man sends me a song that shows up on my disk. Without Apple's actions the copy wouldn't appear. We are both doing it, and Apple has a license to send me a copy. If neither of us had a license, then both would be responsible for the infringement, but that's not the case, there is no infringement because it is an authorized copy.
Yes, the Napster case says that the user downloading a copy is infringing by making a copy. That's because the person uploading the file wasn't authorized to distribute it. If I buy a copy from an authorized distributor, however, who is doing the copying is irrelevant, because it is NOT INFRINGING.
You do have a point that such a tiered service would give an incentive to the carriers to provide poor service so that customers will upgrade - however, I don't see a problem with this kind of "tiering" in general. If I want to pay more to get better service, that's fine. It is the discrimination based on destination or content or application that is the problem.
I have no problem with setting things up so that connecting to a server that pays more to THEIR ISP for better service gives me a faster connection TO THEM, nor do I have a problem with having it faster if I pay more to MY ISP for better service (to everyone, or even to a specific service). The problem is when my ISP wants to charge Google a fee in order to allow ME to connect to Google faster, when Google isn't a customer of theirs, by intentionally degrading the connection if they don't.
Some forms of discrimination are fine: if AOL or Verizon or whoever wants to make a special deal with Google to set up private lines and caching servers to both reduce their costs and improve their customer's connections to Google, and they want Google to share in the cost of that (and presumably it would reduce Google's costs to provide the same level of service), that's fine. That might be a reason to choose one ISP over another. That's a partnership. Distinguishing that from intentionally degrading connections is the difficult part.
Exactly. It doesn't cost more to transfer bits containing video, VOIP, e-mail or web pages. You might need a different QoS for something like VOIP, so charge for special treatment (but don't allow intentionally degrading service just so you CAN charge to not mess it up). Charge by peak bandwidth, delay, jitter, since that's what costs you money - but each individual bit doesn't cost anything. Put in intelligent, fair throttling so everyone gets their share of available bandwidth and you don't need caps. Since it is statistical behavior with a large number of independent actors, it is easy to assign a cost without measuring each individual bit, tracking where it is going, or charging based on what kind of bit it is.
It is only when the bit carriers get into the value-added bits as well that you get a conflict of interest and no longer have an incentive to create an efficient network. That they think they can get away with it demonstrates that they DO have monopoly power, and shouldn't be allowed to do it. If it was truly free market, there would be no problem with them offering a network connection subsidized by a VOIP or music or video service, but the "last mile" issue pretty much forces it to NOT be a free market as long as the carrier is also allowed to offer services other than transport.
I don't misunderstand what a copy is at all. The courts have held that a temporary copy in RAM is "a copy". I am asking Apple to, through their server and their software, to make a copy of a song on my disk. Since they ARE AUTHORIZED to do so, what's the problem?
A band writes a song, performs it, records it, and allows me to order a copy in one of two ways: I can download it or I can have them burn a copy to disk and mail it to me. How is there a difference? When I request that they burn a copy to disk, am I making the copy (since it is my request that causes them to burn it)? The argument is ridiculous. Granted, if it is an UNAUTHORIZED copy, then I don't have any rights to it at all - since I never owned an authorized copy, I can't make any of the the intermediate copies necessary to create a permanent copy on my hard drive. With a song I purchase from iTMS, Apple makes the initial AUTHORIZED copy (which I now own), so there shouldn't be any problem with making intermediate copies to transfer it to my machine, regardless of who initiates it. I don't need a license, Apple does.
Actually, the courts were ruling that you did NOT have a right to make the temporary copies in RAM needed to run a program, which is why copyright law was changed to grant the owner of an authorized copy the right to copy and modify the software to the extent necessary to make it run. This would include copying it from CD to hard drive to RAM, for instance.
If the program won't run unless I click I Agree, and I Do Not Agree with the terms, then patching the program so it works without clicking I Agree might be a "necessary step" to get it to run!
If I reject the license, even if "it said so on the box", I am still the owner of that copy, and can use the content in ways that copyright law allows me.
A terms-of-use agreement on a server is a different matter. You clearly have no right to download content or connect to a server without permission, so if it says that in order to continue you have to agree to certain terms, then those terms ARE enforceable. It is a completely different situation from having an authorized copy of some software in your possession that you own.
The point is, it is an AUTHORIZED COPY. Apple has a license to authorize me to make that copy, if in fact that's what is happening. I don't need a persistent license to play it once I have it, all I need is permission from someone authorized to give it, which is Apple.
I don't understand how you can have it both ways. If Napster was "illegally distributing", then surely a "legal distributor" such as iTMS isn't causing the recipient to make an unauthorized copy!
Look at it from a different standpoint - a reasonable model for selling music in the physical world would be for the store to burn a CD on demand, rather than keeping inventory. If they are authorized to make such copies, and I go in to the store to request they make such a copy, it isn't ME that is making the copy, it's the store. They're the ones authorized to do it, so I don't need a license in order to be allowed to ask them "please, sir, may I have some more?" In the Napster case, SOMEONE was guilty of infringing by making unauthorized copies, and the court rejected the argument that "no, I'm not making copies, he's making copies" by both sides, plus found the man-in-the-middle Napster liable as well. They were all participating in the making of an unauthorized infringing copy. In the case of an authorized copy, you don't need to figure out who actually made the copy, whether it was me by asking Apple's servers to send a copy, or Apple by granting my request, because it is AUTHORIZED.
Apple is making the copy, and they do have a license.
I modify the license to change terms I don't agree with, initial it, then agree to it. Never yet a license that I couldn't agree to after doing that!
If it is a click-license in the software, patch it to change the terms first. Since you haven't yet agreed to the terms that say you can't modify or reverse engineer it, there's no reason you can't do that. Since most software click-licenses show the terms in a standard scrolling text box, you might even be able to make a standard extension to allow you to modify the terms on the screen before you click "I Accept".
Or just patch it so that the buttons for "I Accept" and "I Don't Want To Run The Software That I Paid Good Money For" are reversed. After all, you don't need a license to run it, unless you want to do things that the license allows that you don't already have the right to do.
Generally, it just says "All rights reserved Copyright xxxx Studio Name Here"; it may say "licensed for home use only", but that license must be referring to the license the owner of the copyright gave to the distributor. It isn't a license to me, and I don't need a license.
Copyright law lists certain limited rights that a copyright owner has. If it isn't an explicitly listed right, then the copyright owner can't control it. Non-public "performance" is NOT a right that the copyright owner has control over; that is a right that the owner of the "authorized copy" has (performance for software being "using it"; for a DVD or CD watching and/or listening to it, etc). You don't need a license to read a book, either, but doing a public reading of any substantial portion of it would be a violation of the copyright owner's rights. That doesn't mean you can't read it to your child!
The "unauthorized copying or distribution of this copyrighted ... is illegal ..." messages aren't licenses, nor are they licensing restrictions. They are merely helpful reminders to you of what copyright law says. When it talks about "unauthorized lending", etc, it has little force - it is authorized if copyright law allows it, if copyright law doesn't allow it then it was already an infringement to do that, even if it didn't say so. At best, it could be used to demonstrate that no license was given or implied.
Actually, with DVRs reporting (in a suitably anonymous way) how many people watch a particular ad and how many skip it, they could provide a valuable service to their advertisers, giving them real feedback on what type of advertising actually works. They should embrace the technology, not fight it!
Uh, no? I was responding to your statement that they weren't talking about "direct descendants", but "somehow related" - you said specifically:
I'm disagreeing - if you were related to that person by him being an uncle to one of your direct ancestors, just go back one generation and you have someone who is a direct ancestor! Makes no difference, another 30 years or so.The rest wasn't necessarily disagreeing with you, but expanding on it.
You're right that they're not saying there was an "Adam" 5000 years ago, but they ARE saying that there was a most recent direct descendant to everyone alive today. That doesn't mean that person is the ONLY person alive at the time who had descendants, nor that if you go back one or two more generations, there aren't LOTS of common ancestors, just that one of them has to be the most recent. Going the other way, you will eventually come to a time when EVERYONE then alive was either an ancestor of everyone now alive, or has NO living descendants in this time. Ultimately, you will come down to one individual who is THE common ancestor of everyone alive today, but they aren't saying that was 5000 years ago.
If you have any children, chances are so did your parents.
It is still "prior art", regardless of who used or disclosed it. The inventor gets a year after such disclosure to patent it. Anyone else would have to prove that they invented it first, and in any case lose the ability to patent it after the year is up, even if they did come up with it first.
Point is, there are many ways to implement intermittent wipers; claim 1 just references basically any timing circuit using a capacitor and resistor, which is an obvious way of implementing a timing function, even in 1969. My point about the rain sensor was not whether it could have been implemented back then, but that the basic "idea" of a wiper that could not go just slow or fast, but also intermittently, regardless of the details of how to implement it, should not be patentable - as it was obvious even to a kid back then.
I just don't see the point of being able to patent something that any engineer would come up with when asked how to do something.
I don't think one should be able to get a patent on the idea of an intermittent wiper. A particular implementation, sure. I know that before I ever saw an intermittent wiper, I had the idea of a wiper that would automatically do one wipe based on a sensor that would determine how fast the rain was falling. Doing it manually is actually a step BACKWARD from what I "invented". It's easy to come up with novel ideas, it's the implementation that can be hard, and that's what should be protected.
Prior art doesn't have to be published to invalidate a patent, it only has to be used "in public". That is, if you invent something (say, a novel engine in a car), and you use the invention in public, you can lose the right to patent it a year later even if you haven't revealed the details. A technique used in a program, or even a server, that can be used by people other than the inventors (e.g. people who haven't signed an NDA) can be used as prior art, even if the code implementing the technique is never made public.
This patent was filed in 1998, with a priority claim from an earlier application from Sep. 1997, so anything in public use, or described, or documented, before Sep. 1996 is relevant. In one of TFAs (I think from the other story about Bruce Perens), someone commented that Postgres95 seems to match up with this patent, along with TopLink in Smalltalk in 1994.
I was thinking more about the case where somebody comes back the next day, sees that the bid is well past their original maximum bid point (with several active bidders) and wonders why. You're right that with only one other bidder in an auction, you can't find out how high they're willing to go (which is how they then end up bidding it up a couple bucks at a time repeatedly, trying to find out, but without committing to a higher price - the need to know how much someone else is willing to pay gets them to ignore their own judgement, plus maybe get some sort of vicarious thrill of getting the other guy to pay more for it; in that game, you actually lose when you win the auction).
If 20.53 wins, then 21.00 would have won as well. Ok, so you might save 47 cents. The real problem is, there is no fixed line where it all of a sudden becomes "too expensive". If 21.00 is your maximum, are you SURE you wouldn't buy it for 21.01? It's only another penny. Come on, sir, just one wafer-thin mint? Which grain of sand is the one that changes it from "some sand" to "a pile of sand"? Which strand of hair do you lose that you all of a sudden become "balding"? At which particular second do you become "old"?
It isn't that the bidding is going on for days, but that the item is available for new people to find it and bid on it. The first minute is exactly the same as the last minute, if you have the highest bid.
Sniping won't hurt the rational bidders who put in an early bid at their true maximum price. The only ones who hurt that bidder are the ones who change their concept of how much they're willing to pay based on how much others are bidding (which isn't ENTIRELY irrational; that someone is willing to pay 20% more than you thought it was worth may be an indication that you undervalued it). Sniping avoids THOSE bidders, not the ones who put in their true maximum bid at the beginning, and since there are many auctions where those will be the only bidder, it makes sense to do it.
It works because people often DON'T bid their maximum; they bid a lower amount, and thus if you snipe (and they don't), you get it for less than the maximum they'd actually be willing to pay. The other case is when they THINK they have bid the maximum, but when they see someone else also wants it and has a higher bid, they re-evaluate how much they're willing to pay for it. By sniping, you avoid triggering their irrationality.
However, if eBay wants to allow for sniping, they should simply allow people to put in sealed bids.
You make a good point, but eBay could easily make this easy to deal with. You could make a bid be conditional on being outbid in another auction, create a whole chain of bids, and automatically order them based on auction closing times. They could even handle and resolve cases where multiple auctions close at exactly the same time, and apply first-to-bid rules there as well (among conditional bids only; already-bid non-conditionals would still be "earlier" than any conditional bids, although a conditional bid might still bump up the price at the very end, and thus would have to show up in the bidding history to explain the ending price).
One thing to note, in the commentary in the article, they mention that the study showed an average of fewer than 2 bids per auction. Sure, you're going to win lots of auctions if you bid on items that no one else was interested in (and thus only make one bid), and are much less likely to win an auction if you bid on items that other people are bidding on (especially if you bid and re-bid multiple times, indicating you aren't really sure exactly how much you're willing to pay for it).
eBay could fix this by a) having an auction extended by an hour if any bids come in during the last hour; b) not allowing anyone to bid during that last (extended) hour if they hadn't already bid at least once. This would tend to make people bid on an item earlier, which (since people are bidding irrationally) will tend to drive up prices, which will result in increased fees for eBay.
From the commentary in the USA Today article, they do acknowledge that sniping works because many bidders are naive or irrational. By sniping, you avoid giving these irrational bidders information as to how high you'd go, thus preventing them from overbidding what they would rationally bid.
The other good point in support of sniping is when you're trying to get one item, there are several auctions, and you don't want to commit to any one of them. However, I'd think a better strategy (absent the irrational bidders) would be to bid your maximum on the earliest-to-close auction, then bid your maximum in the next one whenever you get outbid on the first one, and so on. Again, eBay could allow you to make conditional/chained bids like this automatically.
Since sniping is basically just a sealed-bid auction, you're still going to get outbid by an idiot who wants to pay too much. The simple fact that I'm willing to pay more than the current next-highest bidder isn't really giving away a whole lot of information. They still have no idea how high I'm willing to go except to commit to paying more than it is currently. If they're willing to commit to paying more than I am willing, they were going to outbid me with a snipe even if I try to snipe it as well. However, as the article makes clear, what really is happening is that someone sees you want it at a certain price, and that actually changes what they're willing to pay for it.
It doesn't surprise me at all that "the more bids you make on an item, the less likely you are to win it", since you shouldn't ever need to make more than one bid, at the very beginning. If you do get into re-bidding, a) you're irrational and b) you're bidding against someone else irrational, and in that case you certainly are better off avoiding their irrationality and concealing your irrationality by sniping.