You do realize it's illegal to access a computer system without authorization, right? There is a computer equivelant of trespassing, and it actually carries more penalties than normal trespassing.
If you want to argue there's an implict authorization when connecting to an open port, yes, there is, just like there is to walk up to a door and knock on it. But I stuck a sign on the door that said 'come in if you're not selling something', and given my own terms for accessing that computer on that port, and I've clearly printed them out where everyone who connects can see them.
It's not my fault if your software decided to not inform you of these condition. You're in charge of your computer, you shouldn't be using software that hides the fact I'm denying you permission to enter. Your problem, not mine.
What you basically claim is there is no way to tell someone they're illegally in a computer system, and that is clearly legally nonsensical. People have to make a reasonable effort to check if they're trespassing or not. You can't walk up to a door past a sign telling you not to enter and then claim you 'weren't looking that way'.
It's certainly not an excuse to trespass because you hired a guide who stood in front of the sign so you couldn't see it. Your software is an agent of you, and what it does on your command, you do.
The mere fact you continued communications with me shows you got the message. I communicated with you over the only channel possible, and you managed to read the 200- at the start of my message and missed the rest of it. Your problem, not mine.
I've made the notice as large and noticable as I can. And as an additional courtesy, I informed the DMA, who is running a list washing service, I don't want any such email. I have it next to my email address on various web pages. I've made a good faith effort to notify people that I do not give them permission to connect to my machine for the purposes of sending me unsolicted bulk email, and spammers are not making any effort what-so-ever to check if I have or not. They are simply walking onto my computer and using it for whatever purposes they see fit.
And that, folks, is deliberate negligence at the minimum. And negligence on someone's part that leads to a crime is not a defense in any court of law, especially if the 'negligence' suspiciously resulted in profit.
And that makes it trespassing, plain and simple, or the computer equivilent, often referred to as 'hacking'.
No, but it does mean that anyone is welcome to come up to your front porch and knock on your door. If they're selling something, and you don't want it, you can't sue them for trespassing, even though they are standing on your property.
Gee, that's funny, my mail server connection message clearly says that I don't wish to receive any unsolicited email at all. It's literally the same thing as putting a 'No solictitation' sign up, and, yes, you can arrest solictors for trespassing the second they step on your property if you have one of those signs up, no warning required, because you already told them to leave.
Explain to me how it's not trespassing to do the same thing to a mail server.
In your eyes, blah blah blah, whatever, but the Democrats are almost the only people pushing this bill. I think it has one Republican sponsor.
Of course, Leahy is a Democrat, too. But before you try to get a party to fight a bill, maybe you should check if they are the ones pushing it in the first place.
And the Dems have been in Hollywood's pocket much longer than they've been for 'tech issues'. People talk about the 'liberal media', but it's actually the other way around, the 'liberals' are very pro-'media'.
Yeah, because people can't do that already from any random payphone.
Oh, wait, they can.
People are not idiots, and do not use 911 unless they need it. They realize that they need to have a system that works, and it doesn't work when they call 911 for directions to the mall. The remaining few idiots are reminded that they can be fined a rather large amount of money for dialing 911 for inappropriate reasons.
And it would be trivially easy to built in a code that's sent when you dial 911 from these cards which is easily tracable back to the purchaser, which you can't do with pay phones.
While I have no idea whether or not businesses would go for it, this would be insanely useful for 911. Imagine everyone carrying around, in their wallet, next to their credit cards, a card labelled 'emergency' that simply dials 911 when you push a button on it.
Pretty much everyone in the country would spend 10 dollars for one of those things, even if they worked only for 30 minutes due to battery limitations. And you don't have to worry about service, as all cell phone providers are required to carry 911 even for service-less phones.
Nope, under DRM reverse-engineering this copy protection is illegal and importing devices that breach this standard into the US could easily be made illegal. To prove my point how many Russian cars (there are Russian car enthusiasts) are there in California (they totally breach all pollution control laws)?
But the whole point is that people can buy devices to play the music, it's just that computers are 'smarter' and thus can't play the CDs. I fail to see how it would be illegal to turn off error correction in a device, considering other devices are sold without said error correction.
Hmmm I thought mp3s were designed to accept bit flips and shoddy resumes from incorrect files, sorry must have confused it with a WAV file or especially MPEG-L4 designed to aborb these errors.
God no, I get crap MP3s all the time. Though they may be from crap encoders, I don't know.
But in general, obviously, the tighter encoded a a file is lossly encoded, the less errors it can handle. Correcting for errors takes space.
Encapsulating ECC into UDP can be detected by layer 7 routers and blocked.
I don't have any idea what you're talking about, but if you're serious, a simple solution is to just send each packet ten times. But I don't see how a router can the difference between a reply that says 'the last fifty packets received successfully, keep sending' and 'the last fifty packets received, resend packets 32, 33, 35, and 41'.
NASA already blocks packets like this, for >National Security, all packets targetted towards any NASA webservers with port destination set to other than 80 are blocked by all ISPs at the US border (layer 3/4), NO small independent ISPs exist that can send their own packets to the undersea cables bypassing the major telcos.
Now you're talking about firewalling, which doesn't have anything to do with UDP in particular. If companies firewall, people will just go with other ISPs. If it's required they firewall all ports by law, then people will just set up fake FTP connections, which by defination requiring incoming ports, and hence will go though the firewall.
The only way you can keep two computers from talking to each other on the internet is to use NAT, and forward no ports at all, which completely breaks large sections of the internet. It makes all online games unplayable, for one thing.
As for the built-in error correction on core routers - yes, I read about that, but I can't remember if they proliferated much, I can remember PSI.NET had some really advanced routers like that but then they dot-bombed, I thought they were one-of-a-kind, apparently not, don't tempt me to use UDP;-)
The internet doesn't lose packets that much in general, be they TCP or UDP. The place you're most likely to lose data is over a modem, but modems have built-in low-level error correction.
The mostly likely reason to lose data on the wires is bad routing somewhere, where you cannot get to there from here that way, even though it claims you can.
Normal networks, be they fiber optics, ATMs, T3s, simply don't have data disappear en route. Stuff goes out one router, and shows up at the other end of the wires. Only crappy telephone wires have static and crosstalk on them, when people buy more than 50 M/s worth of bandwidth, they will not tolerate lossage. Thus UDP is fairly reliable.
Um, no. OTPs use a different character for each letter. While many of them use a XOR b, each a and b is different.
Using the same value to XOR each letter of the plaintext with doesn't even qualify as 'encryption'. It's just a wacky version of ACSII.
It really amazes me the number of people we have wandering around claim that 'even OTPs are breakable'. No, they aren't, you're just embarassing yourself to claim they are.
You don't have to 'waste' OTP, at least not much. You encrypt the first part of the message, saying something like 'NO MESSAGE', or even a single character that means no message (Starting a message with a period, for example.), with a OTP, then you just dump real random numbers down the line. By 'random numbers', I mean one you haven't gotten around to transferring yet, and hence are not a OTP yet.
This keeps you from using up all your OTP, but is just as secure.
Of course, this is assuming the problems lay in transporting the OTP, not generating the random numbers, but that seems a reasonable assumption.
And, of course, it assumes you're not using the wasteful method of an entire sheet of OTP for a single message, like I believe the Navy uses, at least according to some movies I've seen.
What on earth does 'random' have to do with anything? Is that how you think you decode OTP, you invent random streams of bits till one of the results in a 'plausible' message?
How likely it is that a X length randomly selected string is a real message doesn't change the fact you'd end up with all 'plausible' messages of length X or less as your 'results' to this decryption.
Your arguement makes no sense. ALL combinations are equally likely. You understand that?
It doesn't matter if you know the message is the exact time and place the bombs will ship, you can't tell if it's:
The bombs are coming it at 4:54 on Thursday.
or
The bombs will be here at 1 am this Tuesday.
or
Let's forget the whole bombing and join EFF.
There is literally no way to tell the difference. You can know the entire context of the message, you can know the sender and recipient, you can know every thing except the actual wording of the message, and you still don't know the message.
OTP is unbreakable, period, for exactly this reason. There is no way to tell if you've decrypted the correct message. In fact, you might was well make up your own random string and decrypt it, you'll have just as much luck.
To repeat, when 'decrypting' a OTP, every single character in it could be representing any character, and there is no pattern. With a very improbably OTP, this comment could be the first X characters of the constitution.
I always thought it would be funny to use a non-prime number in asymetrical encryption. (Of the type that multiplies two large primes together, obviously.) Sure, it's theoretically easier to break (If you stumble on any of the factors you've got it.), but what do you want to bet all the brute-forcers automatically skip non-primes? (And while I'm aware that all non-primes are made up of primes, I'm assuming that they don't go around checking any prime that's not roughly the right size.);)
The idea of some NSA goons staring, baffled, as their super computer sits there telling them the keyspace was exhausted and no key was found, and having some math guy wander up and say 'Hey, that number, the one that's supposed to be the product of two 1024 bit primes, is divisable by 113.' is just hilarious.
Even better, hand the person 16 thousand locked boxes, and don't tell him which one is yours.;)
Obscurity is vastly underrated. Having only obscurity is bad, because someone will eventually stumble over the way in, but having a large number of 'apparent' locked doors where only one of them is real is a great addition to any locks you may have on the door. Note the word 'addition'.
However, with encryption algorithms, having them be 'obscure' is having them be 'untested'. And a lot of people know a lot less about encryption than they think they do. (I know just enough to know I don't know anything.)
Keeping an algorithm untested is a great way to ship a product that appears to work, then three weeks later someone comes out with some code that makes your effective keylength be 34 bits, due to mathmatical stupidity.
Solution 1 Obfuscation - Forget CDs, use Read-Only MiniDiscs instead, or any other proprietary format. Problem: it has to be converted to analogue in the earphone (even if this link is encrypted) which can be ripped. This creates extra trouble for the ripper, but for a song to be Napstered only 1 person has to set up the equipment to do this.
If the record companies started distributing songs on minidiscs, within a year people would be buying minidisc players for their computers.
Solution 2 Taint the data - Make the data on the CD different like companies are doing now so that i doesn't meet the CD standard any more but CAN play on most dumb CD players, but not advanced CD players like computers CD drives and so can't be copied, it requires a dongle on the USB port to play.
It really amazes me companies call this 'copy-protection'. It's akin to printing your books with the print sideways so automatic OCRers can't read it.
If any type of 'broken' CD becomes the standard copy-protected version, then CD-ROM drive manufacturers will just start making their drives be able to read it.
Solution 3 Make CD writers illegal. If you really want to copy data or make a backup, then why use CDs which just happen to be compatible with CD drives? Use tape drives instead like Onstream [onstream.com], they're better. So you'll still be able to download mp3s, so what you can only play them on your computer, make mp3 players illegal, people should have proper retail CDs that they carry around.
Making them illegal won't solve any problem at all. People will just come out with something else. And as the real problem is supposed to be people trading music, that won't help at all, as no one trades music with CD-Rs anymore.
Not to mention, it's perfectly legal to trade CD-Rs with people, assuing you use 'music' CD-Rs.
Solution 4 Taint P2P systems - I THINK THIS WOULD BE GREAT! Force music transferred P2P clients to go over UDP, not TCP/IP, thus you get quality degradation in mp3s transferred, same as casette tapes with fair use that the industry asked for. Unfortunately rogue P2P like Freenet could subvert this, forcing CBDTPA to attack Cisco and router manufacturers via ISPs by ordering them to use layer 7 filtering on all traffic to search for mp3 (or whatever) headers. This would signal the end of the free Internet, a very sad day, but the law is *very* powerful
Erm, no. Sorry, doesn't work that way. First of all, if you can 'force' the P2P systems to use UDP, then you can force them to shut down. Second, the only difference between UDP and TCP is that UDP doesn't have the builtin error-correction that TCP does. There's nothing to stop you from adding your own error-correction. Third, UDP does not 'degrade' the signal in the way you're thinking of. A packet either gets there or not, and thus you'd have to invent some new sort of format that works like interlaces GIFs, where you can just get half the packets and you'll get half the song. An MP3 sounds like crap if you drop only 1/1000th of its packets. Fourth, many transmission mediums already have low-level error correction anyway, so UDP is just as reliable as TCP over those links.
Solution 5 Taint the hardware - if none of the measures above works, then this is the nightmare scenario, can you make an x86 compatible processor in your garage? I laugh at all these pathetic people that say some company will not adhere to the standard, what commodity desktop PC processor manufacterer doesn't support big standards e.g. x86/ G3/Sparc, Motorola 68000 or something. Developing an x86 processor on ASIC (otherwise like 1MHz), needs like $100million investment minimum. This provides precision targets for CBDTPA, they will be forced to provide DRM instructions on their processors otherwise they will not be allowed to import to USA, same as heroin. Same with chipset manufacturers, I've yet to see someone make a full-blown Northbridge out of 555 timers and BC108 transistors. They will be forced to provide encrypted tranmission to USB-DRM, IDE-DRM, etc. devices. New DRM drives will be incompatible with non-DRM chipsets and non-DRM processors cannot run on DRM chipsets. This won't necessarily be a difficult transition, the introduction of MMX, SiS motherboards with Northbridge/Southbridge on the same chip, the introduction of DDR hasn't caused any blue smoke and recalls larger than on the scale of 120GXP.
Only inserting a DRM-flagged CD into a DRM drive connected to a DRM chipset with a DRM processor will cause the DRM code in the OS to allow it to play but it would implant an encrypted CPU_ID or DRM_ID into the song every 5 seconds using steganography. The music industry will possess this database, and any music on Napster/Kazaa etc. could be tracked back to source and law enforcement would bust down their door. Solution 1/2/3/4 and CD levies are suddeny starting to sound good now aren't they?
I love the heroin analogy. Because we all know how well that's working. Not to mention the one, fatal flaw...there are other places besides the US. People will just rip it there, and we'll download it here.
And I don't know what on earth you're talking about 'implanting' things into music. If you're talking about doing something to the encryption stream, or any non-music data, we're kind of assuming that the data would be decoded and encoded to MP3 before distribution. If you're talking about watermarking, watermarking simply doesn't work, period. If it's possible to check it, it's possible to remove it. And if the user's
computer is the one adding it, it's even easier to remove it.
Of course, I have yet to see how DRM is supposed to stop someone from setting up a microphone in front of even the most encrypted-all-the-way-to-the-speakers DRM system. (In reality, you'd simply rip open the front of the speakers and attach the 'mic' wires to where the wires attach to the magnets in the speakers, but even literally rerecording using real micphones wouldn't be bad under controlled conditions.)
It's because, all too often, computer specialists end up working for people who don't know anything about computers, but still see fit to dictate rules and regulations about them.
It's akin to you hiring a car mechanic on staff and only telling him to buy parts from Sears. And that all vehicles must use the same viscosity motor oil, and that none of them can have functional cigarette lighters.
And it's fucking annoying. If someone is a specialist, they probably now more than you about what you hired them to do. (Otherwise, you'd be doing it.) Let them do their job. If they say a Linux server will work better, let them us it. If they say open document standards would benefit the company, let them set them up. Etc.
As for Linux on the desktop, it really depends on how much computer experience you have, and, forgive me for saying so, how smart you are. If you've used windows for the last 8 years, and learned it by rote, Linux is a bad idea. If you actually understand Windows and what it's doing, Linux isn't so tricky. If you don't know any OS, Linux is just as easy to learn as Windows, whether you learn it by rote or understanding.
Why the heck do we have all these crazy people saying that X uses Unix domain socket locally? X uses TCP/IP sockets over a network, and shared memory locally. The only reason it opens a socket is so that it knows who to share memory with. I quote from the man pages 'The most efficient local transport will be chosen.'.
Actually...yes. As both replies to this, even the troll, pointed out, what I want is something can replace the functionality of a human being for the specified, sufficiently complicated task. That's exactly right.
And, as an aside, I didn't mean 'grab' my clothes from the hamper with little robotic arms, I just meant that I don't need it to walk around the house. Having an 'dump clothes here' hole in the top is good enough.;) (This is why I must reread my previews for spelling and clarity.;) )
Just because you happen to know stupid Usenet administrators doesn't mean binary newsgroups are bad.
Any sane administrator sets different expiration times for binary and text groups. And, of course, plenty of places don't even carry them.
Perhaps you should point out to your server admin that one day of binary groups equals two weeks of text groups, and instead of having everything at a week expire, he should have binary at six days and text at three weeks?
What gives you the inherent right to an Internet minus spam?
Hey, idiot, read the AUP of your service sometime. Guess what...it's against the AUP to send spam. Wow.
The internet is a private network. The private companies making up the internet have said that they do not wish spam on it. What gives you the right to say otherwise?
Just like you can't rent a hotel room and keep livestock in it, you can't rent an internet conection and send spam over it. Their network, their rules.
Now, there are a few spam friendly ISPs, but, like the man says, we hope they like their intranet, because they sure aren't going to be allowed to connect to the rest of the internet's mail servers.
Talking cars that remember your favorite radio stations, seat settings, A/C settings, and possibly directions to drive to your parent's house are far more likely.
What do you mean, 'likely'. Those things exist right now! We've had programmable radios for a couple of decades, all the new luxury cars have memory on seat settings, and, while I've never seen one, I'm sure the ones with computer maps can remember a few endpoints that you can call up, probably called 'bookmarks' for no apparent reason.
The one thing I haven't heard of are 'remembering' A/C settings, but I have to point out that my car, with mechnical sliders, 'remembers' where I leave them just fine.;) There really isn't any point to having customizable ones, people change those every time they get in the car anyway.
I'm confused, I thought that was the paradox, that you can show that they have to reach the end in a certain amount of time, yet seem incapable of doing so because they have to keep doing more and more movements.
Slashdot Mirror
If you want to argue there's an implict authorization when connecting to an open port, yes, there is, just like there is to walk up to a door and knock on it. But I stuck a sign on the door that said 'come in if you're not selling something', and given my own terms for accessing that computer on that port, and I've clearly printed them out where everyone who connects can see them.
It's not my fault if your software decided to not inform you of these condition. You're in charge of your computer, you shouldn't be using software that hides the fact I'm denying you permission to enter. Your problem, not mine.
What you basically claim is there is no way to tell someone they're illegally in a computer system, and that is clearly legally nonsensical. People have to make a reasonable effort to check if they're trespassing or not. You can't walk up to a door past a sign telling you not to enter and then claim you 'weren't looking that way'.
It's certainly not an excuse to trespass because you hired a guide who stood in front of the sign so you couldn't see it. Your software is an agent of you, and what it does on your command, you do.
The mere fact you continued communications with me shows you got the message. I communicated with you over the only channel possible, and you managed to read the 200- at the start of my message and missed the rest of it. Your problem, not mine.
I've made the notice as large and noticable as I can. And as an additional courtesy, I informed the DMA, who is running a list washing service, I don't want any such email. I have it next to my email address on various web pages. I've made a good faith effort to notify people that I do not give them permission to connect to my machine for the purposes of sending me unsolicted bulk email, and spammers are not making any effort what-so-ever to check if I have or not. They are simply walking onto my computer and using it for whatever purposes they see fit.
And that, folks, is deliberate negligence at the minimum. And negligence on someone's part that leads to a crime is not a defense in any court of law, especially if the 'negligence' suspiciously resulted in profit.
And that makes it trespassing, plain and simple, or the computer equivilent, often referred to as 'hacking'.
Gee, that's funny, my mail server connection message clearly says that I don't wish to receive any unsolicited email at all. It's literally the same thing as putting a 'No solictitation' sign up, and, yes, you can arrest solictors for trespassing the second they step on your property if you have one of those signs up, no warning required, because you already told them to leave.
Explain to me how it's not trespassing to do the same thing to a mail server.
Of course, Leahy is a Democrat, too. But before you try to get a party to fight a bill, maybe you should check if they are the ones pushing it in the first place.
And the Dems have been in Hollywood's pocket much longer than they've been for 'tech issues'. People talk about the 'liberal media', but it's actually the other way around, the 'liberals' are very pro-'media'.
Oh, wait, they can.
People are not idiots, and do not use 911 unless they need it. They realize that they need to have a system that works, and it doesn't work when they call 911 for directions to the mall. The remaining few idiots are reminded that they can be fined a rather large amount of money for dialing 911 for inappropriate reasons.
And it would be trivially easy to built in a code that's sent when you dial 911 from these cards which is easily tracable back to the purchaser, which you can't do with pay phones.
Pretty much everyone in the country would spend 10 dollars for one of those things, even if they worked only for 30 minutes due to battery limitations. And you don't have to worry about service, as all cell phone providers are required to carry 911 even for service-less phones.
But the whole point is that people can buy devices to play the music, it's just that computers are 'smarter' and thus can't play the CDs. I fail to see how it would be illegal to turn off error correction in a device, considering other devices are sold without said error correction.
Hmmm I thought mp3s were designed to accept bit flips and shoddy resumes from incorrect files, sorry must have confused it with a WAV file or especially MPEG-L4 designed to aborb these errors.
God no, I get crap MP3s all the time. Though they may be from crap encoders, I don't know.
But in general, obviously, the tighter encoded a a file is lossly encoded, the less errors it can handle. Correcting for errors takes space.
Encapsulating ECC into UDP can be detected by layer 7 routers and blocked.
I don't have any idea what you're talking about, but if you're serious, a simple solution is to just send each packet ten times. But I don't see how a router can the difference between a reply that says 'the last fifty packets received successfully, keep sending' and 'the last fifty packets received, resend packets 32, 33, 35, and 41'.
NASA already blocks packets like this, for >National Security, all packets targetted towards any NASA webservers with port destination set to other than 80 are blocked by all ISPs at the US border (layer 3/4), NO small independent ISPs exist that can send their own packets to the undersea cables bypassing the major telcos.
Now you're talking about firewalling, which doesn't have anything to do with UDP in particular. If companies firewall, people will just go with other ISPs. If it's required they firewall all ports by law, then people will just set up fake FTP connections, which by defination requiring incoming ports, and hence will go though the firewall.
The only way you can keep two computers from talking to each other on the internet is to use NAT, and forward no ports at all, which completely breaks large sections of the internet. It makes all online games unplayable, for one thing.
As for the built-in error correction on core routers - yes, I read about that, but I can't remember if they proliferated much, I can remember PSI.NET had some really advanced routers like that but then they dot-bombed, I thought they were one-of-a-kind, apparently not, don't tempt me to use UDP ;-)
The internet doesn't lose packets that much in general, be they TCP or UDP. The place you're most likely to lose data is over a modem, but modems have built-in low-level error correction.
The mostly likely reason to lose data on the wires is bad routing somewhere, where you cannot get to there from here that way, even though it claims you can.
Normal networks, be they fiber optics, ATMs, T3s, simply don't have data disappear en route. Stuff goes out one router, and shows up at the other end of the wires. Only crappy telephone wires have static and crosstalk on them, when people buy more than 50 M/s worth of bandwidth, they will not tolerate lossage. Thus UDP is fairly reliable.
All well and good, but that's not what the people were claiming. They were claiming an algorithm for producing random data.
Using the same value to XOR each letter of the plaintext with doesn't even qualify as 'encryption'. It's just a wacky version of ACSII.
It really amazes me the number of people we have wandering around claim that 'even OTPs are breakable'. No, they aren't, you're just embarassing yourself to claim they are.
This keeps you from using up all your OTP, but is just as secure.
Of course, this is assuming the problems lay in transporting the OTP, not generating the random numbers, but that seems a reasonable assumption.
And, of course, it assumes you're not using the wasteful method of an entire sheet of OTP for a single message, like I believe the Navy uses, at least according to some movies I've seen.
How likely it is that a X length randomly selected string is a real message doesn't change the fact you'd end up with all 'plausible' messages of length X or less as your 'results' to this decryption.
It doesn't matter if you know the message is the exact time and place the bombs will ship, you can't tell if it's:
The bombs are coming it at 4:54 on Thursday.
or
The bombs will be here at 1 am this Tuesday.
or
Let's forget the whole bombing and join EFF.
There is literally no way to tell the difference. You can know the entire context of the message, you can know the sender and recipient, you can know every thing except the actual wording of the message, and you still don't know the message.
OTP is unbreakable, period, for exactly this reason. There is no way to tell if you've decrypted the correct message. In fact, you might was well make up your own random string and decrypt it, you'll have just as much luck.
To repeat, when 'decrypting' a OTP, every single character in it could be representing any character, and there is no pattern. With a very improbably OTP, this comment could be the first X characters of the constitution.
The idea of some NSA goons staring, baffled, as their super computer sits there telling them the keyspace was exhausted and no key was found, and having some math guy wander up and say 'Hey, that number, the one that's supposed to be the product of two 1024 bit primes, is divisable by 113.' is just hilarious.
Ironically, the only reason that's illegal is that it used a mechnical device. It's perfectly legal to count cards using your head and fingers. ;)
Obscurity is vastly underrated. Having only obscurity is bad, because someone will eventually stumble over the way in, but having a large number of 'apparent' locked doors where only one of them is real is a great addition to any locks you may have on the door. Note the word 'addition'.
However, with encryption algorithms, having them be 'obscure' is having them be 'untested'. And a lot of people know a lot less about encryption than they think they do. (I know just enough to know I don't know anything.)
Keeping an algorithm untested is a great way to ship a product that appears to work, then three weeks later someone comes out with some code that makes your effective keylength be 34 bits, due to mathmatical stupidity.
Granted, it's hard to decode said key, but it's not 'unbreakable' by any means.
And you've still got the issue of how you transfer the 'selection key' in the first place.
If the record companies started distributing songs on minidiscs, within a year people would be buying minidisc players for their computers.
Solution 2 Taint the data - Make the data on the CD different like companies are doing now so that i doesn't meet the CD standard any more but CAN play on most dumb CD players, but not advanced CD players like computers CD drives and so can't be copied, it requires a dongle on the USB port to play.
It really amazes me companies call this 'copy-protection'. It's akin to printing your books with the print sideways so automatic OCRers can't read it.
If any type of 'broken' CD becomes the standard copy-protected version, then CD-ROM drive manufacturers will just start making their drives be able to read it.
Solution 3 Make CD writers illegal. If you really want to copy data or make a backup, then why use CDs which just happen to be compatible with CD drives? Use tape drives instead like Onstream [onstream.com], they're better. So you'll still be able to download mp3s, so what you can only play them on your computer, make mp3 players illegal, people should have proper retail CDs that they carry around.
Making them illegal won't solve any problem at all. People will just come out with something else. And as the real problem is supposed to be people trading music, that won't help at all, as no one trades music with CD-Rs anymore.
Not to mention, it's perfectly legal to trade CD-Rs with people, assuing you use 'music' CD-Rs.
Solution 4 Taint P2P systems - I THINK THIS WOULD BE GREAT! Force music transferred P2P clients to go over UDP, not TCP/IP, thus you get quality degradation in mp3s transferred, same as casette tapes with fair use that the industry asked for. Unfortunately rogue P2P like Freenet could subvert this, forcing CBDTPA to attack Cisco and router manufacturers via ISPs by ordering them to use layer 7 filtering on all traffic to search for mp3 (or whatever) headers. This would signal the end of the free Internet, a very sad day, but the law is *very* powerful
Erm, no. Sorry, doesn't work that way. First of all, if you can 'force' the P2P systems to use UDP, then you can force them to shut down. Second, the only difference between UDP and TCP is that UDP doesn't have the builtin error-correction that TCP does. There's nothing to stop you from adding your own error-correction. Third, UDP does not 'degrade' the signal in the way you're thinking of. A packet either gets there or not, and thus you'd have to invent some new sort of format that works like interlaces GIFs, where you can just get half the packets and you'll get half the song. An MP3 sounds like crap if you drop only 1/1000th of its packets. Fourth, many transmission mediums already have low-level error correction anyway, so UDP is just as reliable as TCP over those links.
Solution 5 Taint the hardware - if none of the measures above works, then this is the nightmare scenario, can you make an x86 compatible processor in your garage? I laugh at all these pathetic people that say some company will not adhere to the standard, what commodity desktop PC processor manufacterer doesn't support big standards e.g. x86/ G3/Sparc, Motorola 68000 or something. Developing an x86 processor on ASIC (otherwise like 1MHz), needs like $100million investment minimum. This provides precision targets for CBDTPA, they will be forced to provide DRM instructions on their processors otherwise they will not be allowed to import to USA, same as heroin. Same with chipset manufacturers, I've yet to see someone make a full-blown Northbridge out of 555 timers and BC108 transistors. They will be forced to provide encrypted tranmission to USB-DRM, IDE-DRM, etc. devices. New DRM drives will be incompatible with non-DRM chipsets and non-DRM processors cannot run on DRM chipsets. This won't necessarily be a difficult transition, the introduction of MMX, SiS motherboards with Northbridge/Southbridge on the same chip, the introduction of DDR hasn't caused any blue smoke and recalls larger than on the scale of 120GXP.
Only inserting a DRM-flagged CD into a DRM drive connected to a DRM chipset with a DRM processor will cause the DRM code in the OS to allow it to play but it would implant an encrypted CPU_ID or DRM_ID into the song every 5 seconds using steganography. The music industry will possess this database, and any music on Napster/Kazaa etc. could be tracked back to source and law enforcement would bust down their door. Solution 1/2/3/4 and CD levies are suddeny starting to sound good now aren't they?
I love the heroin analogy. Because we all know how well that's working. Not to mention the one, fatal flaw...there are other places besides the US. People will just rip it there, and we'll download it here.
And I don't know what on earth you're talking about 'implanting' things into music. If you're talking about doing something to the encryption stream, or any non-music data, we're kind of assuming that the data would be decoded and encoded to MP3 before distribution. If you're talking about watermarking, watermarking simply doesn't work, period. If it's possible to check it, it's possible to remove it. And if the user's computer is the one adding it, it's even easier to remove it.
Of course, I have yet to see how DRM is supposed to stop someone from setting up a microphone in front of even the most encrypted-all-the-way-to-the-speakers DRM system. (In reality, you'd simply rip open the front of the speakers and attach the 'mic' wires to where the wires attach to the magnets in the speakers, but even literally rerecording using real micphones wouldn't be bad under controlled conditions.)
Which would be, what? 95% of all operations?
It's akin to you hiring a car mechanic on staff and only telling him to buy parts from Sears. And that all vehicles must use the same viscosity motor oil, and that none of them can have functional cigarette lighters.
And it's fucking annoying. If someone is a specialist, they probably now more than you about what you hired them to do. (Otherwise, you'd be doing it.) Let them do their job. If they say a Linux server will work better, let them us it. If they say open document standards would benefit the company, let them set them up. Etc.
As for Linux on the desktop, it really depends on how much computer experience you have, and, forgive me for saying so, how smart you are. If you've used windows for the last 8 years, and learned it by rote, Linux is a bad idea. If you actually understand Windows and what it's doing, Linux isn't so tricky. If you don't know any OS, Linux is just as easy to learn as Windows, whether you learn it by rote or understanding.
Why the heck do we have all these crazy people saying that X uses Unix domain socket locally? X uses TCP/IP sockets over a network, and shared memory locally. The only reason it opens a socket is so that it knows who to share memory with. I quote from the man pages 'The most efficient local transport will be chosen.'.
And, as an aside, I didn't mean 'grab' my clothes from the hamper with little robotic arms, I just meant that I don't need it to walk around the house. Having an 'dump clothes here' hole in the top is good enough. ;) (This is why I must reread my previews for spelling and clarity.;) )
Any sane administrator sets different expiration times for binary and text groups. And, of course, plenty of places don't even carry them.
Perhaps you should point out to your server admin that one day of binary groups equals two weeks of text groups, and instead of having everything at a week expire, he should have binary at six days and text at three weeks?
Vacation request approval, please choose one of the following weeks...
Hey, idiot, read the AUP of your service sometime. Guess what...it's against the AUP to send spam. Wow.
The internet is a private network. The private companies making up the internet have said that they do not wish spam on it. What gives you the right to say otherwise?
Just like you can't rent a hotel room and keep livestock in it, you can't rent an internet conection and send spam over it. Their network, their rules.
Now, there are a few spam friendly ISPs, but, like the man says, we hope they like their intranet, because they sure aren't going to be allowed to connect to the rest of the internet's mail servers.
What do you mean, 'likely'. Those things exist right now! We've had programmable radios for a couple of decades, all the new luxury cars have memory on seat settings, and, while I've never seen one, I'm sure the ones with computer maps can remember a few endpoints that you can call up, probably called 'bookmarks' for no apparent reason.
The one thing I haven't heard of are 'remembering' A/C settings, but I have to point out that my car, with mechnical sliders, 'remembers' where I leave them just fine. ;) There really isn't any point to having customizable ones, people change those every time they get in the car anyway.
I'm confused, I thought that was the paradox, that you can show that they have to reach the end in a certain amount of time, yet seem incapable of doing so because they have to keep doing more and more movements.