actually, why not form a correlation? If the opposite occured (i.e., napster came out during a recession and was shut down at the beginning of a recovery), the record labels wouldn't hesitate to use it in their favor
This reminds me of a little story (fictional, I hope)...
A company that produces traffic light control systems dicovered that they had a bug in their logic. You see, when a pedistrian would approch the intersection, and push the little button on the light post to activate the walk signal, all would work as expected... however, if the button was held for more than 10 seconds, the traffic signal would turn green in both directions, thereby causing a hazzardous traffic situation. Therefore, the company spent several million dollars on campain contributions and lobying efforts in order to get a law passed that made it illegal to press the button in this manner on this brand of traffic signal. They also had installed fingerprint scanners on all the buttons so that evidence could be collected to arrest and convict anyone who wanted to cross the street.
From what it seems, there's a bit more to it than this. I would immagine that each version of the cd has a different watermark on it. They know that it is still gonna be ripped & coppied, they just want to find out which protection scheme is gonna be broken by the most people. The version with the least number of copies out there will be the format the industry settles on.
The problem is, instead of Microsoft going after the developers, they will simply go after any business user that uses the software. Just because something is legal where it's made, doesn't mean it is legal to import and use it here...
The best one-time password system i've used is a SecurID card (Security Dynamics, which is now owned by RSA security), it displays a different 6-digit number every minute, you use it plus a pin code as a second-level password.
Another solution if you're on a budget is to have a list of several hundred random numbers with a key value associated with them, recorded in a file on your machine, and carry a printout of them folded up in your wallet. Have your authentication software ask you for one of the numbers at random...
Actually, I remember GM getting sued for the Chevy Beretta, since there is a firearms company by that name. Of course, in this case, there isn't a generic word "beretta" (except maybe "biretta", from the Italian "berretta", meaning a clergymen's cap), so the firearms company could claim that it could be mistaken that they were endorsing the car. It turned out in the court case that GM ended up paying the firearms company a whole sh*tload of money.
That would be a good start, but we also need the driver<>os and driver<>hardware api well defined, something similar to Intel's UDI (uniform driver interface). Too bad there was so much opposition to it (mostly for political reasons, not technical, i.e., "we don't want windows users to be able to take advantage of Free Linux drivers", and "this will only encourage the proliferation of binary drivers"). Now don't get me wrong, I perfer open source/Free drivers to binary only, but if a binary driver is the only option, I would rather it be firewalled off in an api, ideally so that it runs in user space and communicates to an open part running in kernel mode via the api. I'm not sure if that's how UDI was supposed to work, but I want something like this. Also, it would be nice if a driver that is open source but not in the mainline kernel didn't need to be updated everytime a new kernel is released. Something similar to UDI would help in this area. BTW, does anyone know if the UDI spec was technically elegent? Is it fesible to come up with something similar that is done wright? (at least for drivers for non-critical hardware... I still want my scsi driver to be part of the kernel).
The most likely way for this to happen is for manufacturers to only give you the option of leasing a computer instead of buying. However, as long as there is a market sold vs. leased computers, there will always be products available.
Another possibility is for the market for "open" components to be artificially diminished. Lets say that some new operating system (which becomes very popular, with 90% market share) requires a custom licensed chip for copy protection. Since this new OS will not run on open systems that don't have this proprietary chip, there will be little deminde for the open hardware, the market dries up,...
However, the most likely threat is a trend in hardware that has been going on the last few years. As processors get faster, there are more spare cpu cycles for the average applicatoin. Also, since it is cheaper to manufacture software than it is hardware, more devices have come on the market with much of the hardware functionality replaced by software (winmodems, winprinters, winsoundcards, etc.), with the software component of thesed devices are usually only written for the most popular OS's. The trend for these types of devices will only accelerate as cpu power becomes more abundant and demand for more expensive non-software based components drops off. The only solution I can see for this situation is to develop a mini-os that runs inside a virtual machine inside your main os, with a well defined api. The api will talk to the parent os, which in turn talks to the hardware. Thie mini-os should be simple enough that it can be implimented under any system, while being capable of running the same binary drivers no matter which parent os you're running.
Actually, your car analogy isn't quite right. It would be more like... you hire me to pick up your car at your house and take it to have it washed. On my way to the car wash, I take a detour to a McDonalds Drive through, then I pick up a date to go cruising, then get the car washed and return it to you (with a few more miles and wear & tear). You couldn't prosecute me for stealing your car, since I had authorization and was hired to drive it, but you could potentialy sue me for the extra milage.
Another car analogy would be a limo or cab driver that takes the uses the car for personal business during off hours.
I got automatically switched to Telocity when my previous ISP, PhoenixDSL, was bought out. Since I had IDSL, I didn't get Telocity's fancy gateway, they told me to continue to use my old 3com IDSL box (which I owned). I was with Telocity for 2 weeks when Northpoint died, and Telocity told me there was no other IDSL providor at my location (actually, there was, but none that would sell the service for 49.95/mth). I have since switched to a Rhythmns business idsl line through a different ISP, but kept getting my credit card billed by Telocity (of course, I disputed each charge with my credit card company, so I didn't have to pay them). Also, Telocity keeps contacting me wanting their gateway back (remember, my idsl line didn't get their gateway, so I don't have it to give back to them). I have since canceled the credit card I used with them to keep from getting a supprise charge, and hadn't heard from them for three months now.
I agree that it is usually a problem with the application, not the library, that causes this issue. Sorry that I don't have specific examples, but I remember having all kinds of problems when I ran a slightly out of date distribution, and when glibc 2 came on the scene. There were bugs in one minor version of glibc2, and the next minor released fixed those bugs but ended up breaking existing code. I remember at the time having virtually no success trying out various new programs without them seg-faulting, unless I updated various libraries, which in turn would cause other programs that did work to stop working and start segfaulting. I eventually solved most of my problems by deciding to keep the latest version of several distributions installed, then I eventually switched over to FreeBSD and would wait for someone to relase a PORT skeleton for that program. This worked because if a port was written for FreeBSD 3.2, for example, it would generally work for everyone's FreeBSD 3.2. (Not saying that FreeBSD's port system isn't perfect, it doesn't address updating ports easily). When I get time, I'm gonna switch over to Debian (I had Corel Linux before), it appears that apt-get address most of these issues.
>..., there is a little something called library versioning...
This still has one major drawback:
libsomething.so is a symlink to libsomething.so.2, which in turn points to libsomething.so.2.17. The problem is when an app was coded and debuged on a system with version 2.17, but the end user only has 2.15. In theory, there shouldn't be any incompatible changes between 2.15 and 2.17, so the user just needs to load up to 2.17 and point.so.2 to.so.2.17. But in practice, in many situations this will cause existing binaries that point to.2 (which now points to 2.17) to break. So, the only solution is to load.so.2.17 into a new directory, and set LD_LIBRARY_PATH or LD_PRELOAD accordingly before starting up that app.
Actually, there is a difference... not in quality, but the Audio CDR's have a code on them that says it's an audio cd. What this does is it allows dedicated cd recorders (audio, consumer grade, no pc involved) to reject non-audio cdr's. I have also heard rumor that some newer cd players will reject data cdr's, but havn't seen proof yet.
Well, a properly implimented Thought to Computer (tm) interface would function similar to how our thoughts control various parts of the body. Just because you "think" about your hand with a knife in it plunging into that bully that tourmented you through highschool doesn't actually cause your hand to perform that action. It's a seperate process.
The only problem is, an adult's brain wouldn't know how to interact with a new set of input/output neurons. The only way this would be feasable would be to implant this into the brain while it is still developing.
...instead of having the device spit out "encrypted" (actually, encoded) content, they should have desined their firmware to only scan a barcode that begins with a particular sequence, then require that anyone (such as Radio Shack catalogs) print their bar codes beginning with that sequence. This would guarantee that no one would use the cuecat as a free generic bar-code scanner
It can also be said that the only reason for running anon ftp or http is to share files to the world. However, there are reasons for shareing your hard drive/folder other than for world access. I.e., if a user had a home network before getting a cable modem/dsl, and shared a desktop drive for access from a laptop, then a year later got cable modem, forgetting that there were open shares....
The only way I can see this being kept "pure" is for the court's experts to not have knowledge of pending cases. Instead, judges would query the experts to explain technological issues to them, and to educate them on how things work, without revealing the nature of the case. Therefore, the experts used would not be able to (voluntarily/involuntarily) insert personal bias.
Actually, I think the main benefits of mining asteroids is not for shipment of material back to earth, but to refine and manufacture parts for large spacecraft that would otherwise be to expensive to launch into space from earth.
This should be enough to get the patent overturned, but the reason why the patent was granted to begin with is that the patent examiners don't go on a massive hunt for prior art. They typically look in areas such as industry trade journals (such as Dr. Dobbs, I would assume). Also, just because a piece of code is released doesn't necissarily constitute prior published art, unless the algorithm is documented. Of course, this only relates to sufficient prior publication to prevent a patent from being granted. Published prior source code may be considered enough to overturn the patent (although a published binary without source probably wouldn't).
It basically boils down to that the patent office just checks to make sure the patent is written up in the proper legal syntax, if so then they grant it and let the courts figure out if it has merrit.
Basicaly, it's like this: A patent examiner has a limited amount of time to process a patent. The patent office gets money for each patent they grant (whether or not it's overturned later). The examiner only looks for published prior art (such as trade journals). They typically can't refuse a patent based on the "obvious to one skilled in the art" clause (unless it is something blatently obvious), because that would constitute the individual examiner making a judgement call based on her personal opinion... there's just too much room for inconsistencies among the various examiners. Therefore, they let the courts decide whether the patent has merrit.
Also, just because someone has used a technique before, it only constitutes prior art if the technique was published. If the technique is kept secret (i.e., closed source), then it is considered a "trade secret", in which another party is free to discover the technique on her own (and subsequently patent the discovery). If I recall correctly, it used to be the case that if someone got a patent on something you held as a trade secret, they could come after you and charge royalties for using their patent, even though you've been using the "invention" for a number of years. This clause has been recently cleared up, so that any previous inventors of a patented invention can continue to use said invention in their own products, they just can't license it out to other parties.
This is pretty simple. A person out of work is generally considered a burden on society (unless said person has enough personal resources...) Therefore, most states have laws that say a person can get in big trouble if said person intentionally sabotages another's job. For example, if a prospective employeer calls one's previous employeer for a refference, the most the previous employeer can say is "Yes, Jack worked here, from this date to this date". If they state any opinions such as "Jack was a slackoff, lazy good for nuttin'...", then Jack can sue said former employeer. At least thats the way it is in my state.
Actually, the reason CO is dangerous is because it because it binds more tightly to the hemogloben (sp?) (the sticky part of red blood cells) than does oxygen, thereby rendering those blood cells useless. CO2, otoh, binds less tightly than oxygen, which is what allows our lungs to exchange the co2 for oxygen...
Slashdot Mirror
actually, why not form a correlation? If the opposite occured (i.e., napster came out during a recession and was shut down at the beginning of a recovery), the record labels wouldn't hesitate to use it in their favor
A company that produces traffic light control systems dicovered that they had a bug in their logic. You see, when a pedistrian would approch the intersection, and push the little button on the light post to activate the walk signal, all would work as expected... however, if the button was held for more than 10 seconds, the traffic signal would turn green in both directions, thereby causing a hazzardous traffic situation. Therefore, the company spent several million dollars on campain contributions and lobying efforts in order to get a law passed that made it illegal to press the button in this manner on this brand of traffic signal. They also had installed fingerprint scanners on all the buttons so that evidence could be collected to arrest and convict anyone who wanted to cross the street.
From what it seems, there's a bit more to it than this. I would immagine that each version of the cd has a different watermark on it. They know that it is still gonna be ripped & coppied, they just want to find out which protection scheme is gonna be broken by the most people. The version with the least number of copies out there will be the format the industry settles on.
The problem is, instead of Microsoft going after the developers, they will simply go after any business user that uses the software. Just because something is legal where it's made, doesn't mean it is legal to import and use it here...
Another solution if you're on a budget is to have a list of several hundred random numbers with a key value associated with them, recorded in a file on your machine, and carry a printout of them folded up in your wallet. Have your authentication software ask you for one of the numbers at random...
Or, better yet, have a remote turn-on... swallow the thing, and a few hours later when it is positioned properly, send it a radio signal to activate.
Actually, I remember GM getting sued for the Chevy Beretta, since there is a firearms company by that name. Of course, in this case, there isn't a generic word "beretta" (except maybe "biretta", from the Italian "berretta", meaning a clergymen's cap), so the firearms company could claim that it could be mistaken that they were endorsing the car. It turned out in the court case that GM ended up paying the firearms company a whole sh*tload of money.
That would be a good start, but we also need the driver<>os and driver<>hardware api well defined, something similar to Intel's UDI (uniform driver interface). Too bad there was so much opposition to it (mostly for political reasons, not technical, i.e., "we don't want windows users to be able to take advantage of Free Linux drivers", and "this will only encourage the proliferation of binary drivers"). Now don't get me wrong, I perfer open source/Free drivers to binary only, but if a binary driver is the only option, I would rather it be firewalled off in an api, ideally so that it runs in user space and communicates to an open part running in kernel mode via the api. I'm not sure if that's how UDI was supposed to work, but I want something like this. Also, it would be nice if a driver that is open source but not in the mainline kernel didn't need to be updated everytime a new kernel is released. Something similar to UDI would help in this area. BTW, does anyone know if the UDI spec was technically elegent? Is it fesible to come up with something similar that is done wright? (at least for drivers for non-critical hardware... I still want my scsi driver to be part of the kernel).
Another possibility is for the market for "open" components to be artificially diminished. Lets say that some new operating system (which becomes very popular, with 90% market share) requires a custom licensed chip for copy protection. Since this new OS will not run on open systems that don't have this proprietary chip, there will be little deminde for the open hardware, the market dries up, ...
However, the most likely threat is a trend in hardware that has been going on the last few years. As processors get faster, there are more spare cpu cycles for the average applicatoin. Also, since it is cheaper to manufacture software than it is hardware, more devices have come on the market with much of the hardware functionality replaced by software (winmodems, winprinters, winsoundcards, etc.), with the software component of thesed devices are usually only written for the most popular OS's. The trend for these types of devices will only accelerate as cpu power becomes more abundant and demand for more expensive non-software based components drops off. The only solution I can see for this situation is to develop a mini-os that runs inside a virtual machine inside your main os, with a well defined api. The api will talk to the parent os, which in turn talks to the hardware. Thie mini-os should be simple enough that it can be implimented under any system, while being capable of running the same binary drivers no matter which parent os you're running.
Another car analogy would be a limo or cab driver that takes the uses the car for personal business during off hours.
I got automatically switched to Telocity when my previous ISP, PhoenixDSL, was bought out. Since I had IDSL, I didn't get Telocity's fancy gateway, they told me to continue to use my old 3com IDSL box (which I owned). I was with Telocity for 2 weeks when Northpoint died, and Telocity told me there was no other IDSL providor at my location (actually, there was, but none that would sell the service for 49.95/mth). I have since switched to a Rhythmns business idsl line through a different ISP, but kept getting my credit card billed by Telocity (of course, I disputed each charge with my credit card company, so I didn't have to pay them). Also, Telocity keeps contacting me wanting their gateway back (remember, my idsl line didn't get their gateway, so I don't have it to give back to them). I have since canceled the credit card I used with them to keep from getting a supprise charge, and hadn't heard from them for three months now.
I agree that it is usually a problem with the application, not the library, that causes this issue. Sorry that I don't have specific examples, but I remember having all kinds of problems when I ran a slightly out of date distribution, and when glibc 2 came on the scene. There were bugs in one minor version of glibc2, and the next minor released fixed those bugs but ended up breaking existing code. I remember at the time having virtually no success trying out various new programs without them seg-faulting, unless I updated various libraries, which in turn would cause other programs that did work to stop working and start segfaulting. I eventually solved most of my problems by deciding to keep the latest version of several distributions installed, then I eventually switched over to FreeBSD and would wait for someone to relase a PORT skeleton for that program. This worked because if a port was written for FreeBSD 3.2, for example, it would generally work for everyone's FreeBSD 3.2. (Not saying that FreeBSD's port system isn't perfect, it doesn't address updating ports easily). When I get time, I'm gonna switch over to Debian (I had Corel Linux before), it appears that apt-get address most of these issues.
This still has one major drawback: .so.2 to .so.2.17. But in practice, in many situations this will cause existing binaries that point to .2 (which now points to 2.17) to break. So, the only solution is to load .so.2.17 into a new directory, and set LD_LIBRARY_PATH or LD_PRELOAD accordingly before starting up that app.
libsomething.so is a symlink to libsomething.so.2, which in turn points to libsomething.so.2.17. The problem is when an app was coded and debuged on a system with version 2.17, but the end user only has 2.15. In theory, there shouldn't be any incompatible changes between 2.15 and 2.17, so the user just needs to load up to 2.17 and point
Come on. Thats like saying that programmers purposly put bugs in their software or leave out features just so they can sell you future upgrades.
Actually, there is a difference... not in quality, but the Audio CDR's have a code on them that says it's an audio cd. What this does is it allows dedicated cd recorders (audio, consumer grade, no pc involved) to reject non-audio cdr's. I have also heard rumor that some newer cd players will reject data cdr's, but havn't seen proof yet.
The only problem is, an adult's brain wouldn't know how to interact with a new set of input/output neurons. The only way this would be feasable would be to implant this into the brain while it is still developing.
...instead of having the device spit out "encrypted" (actually, encoded) content, they should have desined their firmware to only scan a barcode that begins with a particular sequence, then require that anyone (such as Radio Shack catalogs) print their bar codes beginning with that sequence. This would guarantee that no one would use the cuecat as a free generic bar-code scanner
This won't be possible until Amazon's patent is overturned.
It can also be said that the only reason for running anon ftp or http is to share files to the world. However, there are reasons for shareing your hard drive/folder other than for world access. I.e., if a user had a home network before getting a cable modem/dsl, and shared a desktop drive for access from a laptop, then a year later got cable modem, forgetting that there were open shares....
The only way I can see this being kept "pure" is for the court's experts to not have knowledge of pending cases. Instead, judges would query the experts to explain technological issues to them, and to educate them on how things work, without revealing the nature of the case. Therefore, the experts used would not be able to (voluntarily/involuntarily) insert personal bias.
Actually, I think the main benefits of mining asteroids is not for shipment of material back to earth, but to refine and manufacture parts for large spacecraft that would otherwise be to expensive to launch into space from earth.
It basically boils down to that the patent office just checks to make sure the patent is written up in the proper legal syntax, if so then they grant it and let the courts figure out if it has merrit.
Also, just because someone has used a technique before, it only constitutes prior art if the technique was published. If the technique is kept secret (i.e., closed source), then it is considered a "trade secret", in which another party is free to discover the technique on her own (and subsequently patent the discovery). If I recall correctly, it used to be the case that if someone got a patent on something you held as a trade secret, they could come after you and charge royalties for using their patent, even though you've been using the "invention" for a number of years. This clause has been recently cleared up, so that any previous inventors of a patented invention can continue to use said invention in their own products, they just can't license it out to other parties.
This is pretty simple. A person out of work is generally considered a burden on society (unless said person has enough personal resources...) Therefore, most states have laws that say a person can get in big trouble if said person intentionally sabotages another's job. For example, if a prospective employeer calls one's previous employeer for a refference, the most the previous employeer can say is "Yes, Jack worked here, from this date to this date". If they state any opinions such as "Jack was a slackoff, lazy good for nuttin'...", then Jack can sue said former employeer. At least thats the way it is in my state.
Actually, the reason CO is dangerous is because it because it binds more tightly to the hemogloben (sp?) (the sticky part of red blood cells) than does oxygen, thereby rendering those blood cells useless. CO2, otoh, binds less tightly than oxygen, which is what allows our lungs to exchange the co2 for oxygen...