This is indeed *exactly* what they should do. The real trick is that the watermark-decoding should be kept completely secret. If there is any way to detect a watermark then a pirate can easily figure out whether they have removed it or not and the whole thing is worthless. Without any detector program the pirate can never know if they will be tracked down. Unfortunatly I can see a huge temptation to make a player that refuses to play "unauthorized" content, which will wreck the whole scheme. But maybe they are smart enough to resist this.
Besides idiot who will try to make the "won't play unauthorized content player", you mention the real threat to this. The *AA companies are using piracy as an excuse for DRM. They certainly don't want to see piracy stopped as this will remove their argument for DRM. DRM means they can enforce pay-per-view, and eventually (once enough schemes are broken that they can convince the government to make players of non-encrypted content illegal) they can set up a system where "unlicensed" content producers (ie amateurs) cannot distribute entertainment, as distribution requires purchase of an expensive encryption license.
Please everybody, support this as much as possible. This is the real solution to piracy.
Basically, why did they even add the option to their DRM to prevent "squirting"? Did the seriously think that the paranoid recording companies would *not* use that option if it was there? If they had the slightest brains at all, they would have realized that they had in fact disabled one of the few features of the Zune. And they certainly should have known they had the power to refuse to implement such an option. Most of the DRM is there to lock people into using the Zune and thus satisfies Microsoft's own personal interests, but this option makes no sense what so ever!
No matter whether you think Microsoft is Satan's spawn or the greatest company in the world, it is obvious this decision is completely against their desires. This is proof positive that they are a bunch of idiots. This is scary yet relieving at the same time...
Just to be more accurate, the thing that Microsoft really made sure was impossible was dual-boot. They knew that that could lead to alternative os's. At the time I fully expected machines would be sold with Windows plus a "game operating system" and all games would require you to reboot to the game system. Microsoft realized this as well and made absolutely sure that such a machine would never be produced.
Machines that just run an alternative system were never as much of a problem for Microsoft, and today they allow them. They rely on the fact that you need Windows to run Word or Quicken or TurboTax, so any machine that does not run Windows is worthless. So even today they make sure it is impossible to sell a dual-boot machine, while they allow alternative systems so they can pretend they are obeying the monopoly rules.
must be possible to convert existing Word documents to it without any loss. In order to do this, there must be a one-to-one mapping between the.DOC semantics and the OpenXML semantics.
Uh, no, that is wrong. A many-to-one mapping would work for that requirement. If doc has many ways of representing the same information, turning them all into the same way would work just fine and would be lossless.
A requirement that it lossly convert *both* ways, however, would require this. Otherwise converting to XML and back would probably map the many ways of representing the same data to a single one and would be lossy. However I am willing to bet that OOXML does not satisfy that criteria. So this excuse is totally bogus.
Take a look above and you will find an equally-irate person complaining that the Democrats got full names and that they thought this was disparaging to the Democrats. This was somewhat backed up by responses which indicated that the original submission had the title "Democrats push new DRM bill" which certainly implies the original wording was done by somebody who did not like Democrats.
The should fix these articles by naming everybody and putting (D) and (R) after each name, so you and that guy will stop throwing hissy fits. Apparently any discrepency in how things are decided is considered an insult by both sides...
Uploaders don't want to hear this, the hardware manufactures don't want to hear it, and Microsoft *certainly* does not want to hear this. But the content guys, if they want their DRM to work, have to stop listening to Microsoft's promises and require the decoding to be done in *hardware*.
A more complex requirement, one that is going to be difficult to explain, is that the hardware API must be completely documented and exposed, so that a Linux driver is easy to create. Obfuscation never works, while exposing everything will mean that any mistakes will be pointed out instantly by hackers trying to get fame by showing how smart they are. Being secretive about the card is *proof* that the design is not sufficently safe and the content industry should not license it. It is also pretty obvious that about 90% of the work in breaking DRM is by people trying to play them on Linux (real pirates (not uploaders) are much more interested in copying the disk image including the DRM, not in decoding it), so this would greatly reduce the number of smart people trying to crack it. Note that a custom driver will be pretty much equivalent to building your own IR remote for a dvd player, it won't do a lot.
My guess is the card will actually be inline between the graphics card and the display. It would replace a keyed area with the video and force the HDCP on and send it out the cable to the display. It probably also needs to directly connect to the disk player so that a disk image could not be fed to it. There may be schemes of key exchange so that such hardware connections are not needed (software only has access to encrypted and un-reusable data). The system api would probably be pretty much the same as the buttons on a remote control.
You might want to check your reading comprehension.
He was talking about putting music purchased from a site ONTO a cd. Not taking music from a cd and putting it on the computer/player. Apple allows a standard non-DRM CD to be made from iTunes AAC files.
You want to pretend that Slashdot is closed minded, but it appears you are. Did you actually read the comments, or are you just making assumptions about what they said? I would say you are making assumptions.
I did read the comments and I would say that virtually 100% accuse this guy of spreading FUD and wanting to be able to install spyware. About the only negative comments on MicroSoft were about them trying to lock games into their platform and Vista, which has nothing to do with what the original article was about, he absolutly does not care about the lock in.
I think you are missing the fact that there was *already* a (non GPL) Windows port from TrollTech. This kind of discouraged anybody from porting the X version to Windows. Some people who would want to just bought the Windows version. Others quite likely decided that a fork like that would be a bad idea, since the two Windows versions were unlikely to be compatable. A lot of possible programmers did not like the idea of doing the work to write something that already existed. There may also have been fear that somebody would insert code from the Troll Tech version (you could get the code easily even though it is not GPL) and violate copyrights.
I suspect that if Qt did *not* have a Windows port but just a GPL X version, there would be a GPL Windows version already, probably several years ago.
No, I would pop up a warning that says "SOME FORMATTING MAY BE LOST", just like Microsoft does with EVERY other format it can save to (whether or not anything will be lost). Now maybe, having access to the source code, I would try to be nice and CONVERT (look up the word "convert" if it is foreign to you) to the new format, if it was possible. But that is not important. I CERTAINLY would not put a marker in the program that says "hey, YOU convert for me, as I don't want to!".
What they are doing is indefensible, so don't even try.
The problem with that solution is that it will allow software other than MSWord to read the saved documents correctly by making the spec simple enough to implement. Though not in the ISO submitted documentation, "make sure it is impossible for other software to read this correctly" was one of the primary design criteria, so the solution you propose was eliminated immediately.
No, MSWord should pop up a scary-sounding warning like "SOME FORMATTING MAY BE LOST" and save the file the best it can.
Rest assurred that if they implemented ODF it would pop up the most scary warnings possible, saying the user's very life may be threatened if they save this way, even if nothing would be lost. They should apply the same rules to their own code, and stop this bullshit.
Searching UTF-8 works by using string subsearch, which has the advantage that it can also search for any word, not just some glyph. Also you should check real code, I very much doubt anybody searches for any single unicode character other than some ASCII punctuation, which could be searched for in UTF-8 by the normal ASCII search (since it won't match any bytes of a multibyte character).
UCS-4 (the correct name for UTF-32 because no actual transformation is done) is useful, but mostly as for graphics when actually drawing the glyphs. I recommend that all storage be done in UTF-8, mostly because it is extremely difficult to eliminate all ASCII internally, so this is the only way to make all your string data be the same.
The glib people apparently though UCS-4 would be useful, that is why wchar_t on Linux is 4 bytes, instead of the 2 used by everybody else, including other Unix implementations. However the main effect is that this means code using wchar_t is incompatable between Windows and Linux, and that nobody on Linux uses it. About the only positive effect I see is that it discourages wchar_t usage.
I believe Unicode does have problems, but "surrogate pairs" is strictly a problem with the UTF-16 encoding chosen by Microsoft (and by the Unix companies before that) because they failed to realize the impossibility of having fixed-size characters. Plan9 chose UTF-8. Linux pretty much chose it as well, I would like to think it is because they were smart, but the main reason is that it is enoumously easier to implement UTF-8 on an existing system than to add "wide characters".
Unicode's problems are chiefly the insistence that all existing character sets have a lossless translation from/to Unicode. This is why there are hundreds of combined character+accent mark glyphs, when it would be a lot easier for everybody if only decomposed characters were supported. I think there may be similar problems with all the pictrograph characters for Chinese and Japanese, in that they could instead be "composed" of several "stroke" characters, but I certainly can't say anything about languages I know nothing about.
A good deal of problems with Unicode and encodings I think stem from a politically-correct view that it is somehow bad if the Americans get the "better" short characters. Thus the insistence that things all be equal-sized when stored and that characters be composed.
You are apparently totally ignorant of modern design if you think storing strings as bytes preculdes Unicode. There is something called UTF-8. Look it up sometimes. You might need to hit yourself with the cluestick a few times to remove the delusion that there is some advantage in having all the glyphs take the same number of bits and/or that there is any solution out there where the glyphs do all take the same number of bits. Hints: 1.Text is made of WORDS, sentences, paragraphs, lines, and many many other structures that are variable length. 2.There are things called "surrogate pairs" in utf-16. 3.There are "combining characters".
The fact that D jettisoned a whole lot of crap for "wide characters" is one of the best indications that they get it and know what they are doing!
The problem is not coming up with a question that a computer program cannot answer. The problem is making a computer program that can create such questions.
If the questions are created by a human, there is going to be a limited set. The spammers only have to figure out the answers to that limited set. Only by having the computer generate a (essentially) infinite set of questions can this workaround be avoided.
Anybody who trusts security through obscurity is an idiot, and you have just shown you are one.
Any proper design MUST assume the bad guys have every single bit of information about the machine. Open source software may be a way to make sure the designers do not miss this assumption, and thus is extremely good for security of the machine.
No, that is a lie that is actually being propagated by conservatives (!) to stop investigation of electronic voting, by making people think it will upset the entire country and threaten the conservative leadership.
In fact the last election would be an *excellent* time to investigate the electronic voting, for precisely the reason that it *can't* affect the outcome. Even if you assumme every single electronic machine in Ohio was fixed to throw the election to the Republicans, it would require the counties with those machines to really be voting something like 70% for Kerry, while adjacent and otherwise identical counties without electronic machines somehow were voting 49% for Kerry, for the election to come out differently.
Of course the conservatives are scared to death that something funny did happen and will be detected, so they will make sure any attempt to detect it is avoided by a bogus threat that it will destabilize the country. And a bunch of nutjob Liberals who refuse to admit that perhaps the majority in the country don't agree with them are doing there best to help this plot along...
BeOS was of course killed by Microsoft refusing to allow OEM's to sell dual-boot machines.
I'm sure lots of other stuff was killed without even seeing the light of day because of this. Just before BeOS, I fully expected machines that would run Windows and that you rebooted in order to play games into a special game OS, which would allow another company making that system to compete, and eventually the Game OS would become so powerful that Microsoft would have real competetion. But Microsoft stopped that dead by not allowing dual-boot. The only reason you heard of BeOS is that they thought their system was so good that maybe they could also sell non-dual-boot machines.
Plenty of people have asked where the preinstalled Linux machines are. Well they somewhat exist, businesses buy them. But just check what *everybody* using desktop Linux really has, and you will find that 100% of them are dual-boot machines. Microsoft is not stupid, they know that dual-boot is the real killer and will do anything they can to stop it.
Slashdot Mirror
This is indeed *exactly* what they should do. The real trick is that the watermark-decoding should be kept completely secret. If there is any way to detect a watermark then a pirate can easily figure out whether they have removed it or not and the whole thing is worthless. Without any detector program the pirate can never know if they will be tracked down. Unfortunatly I can see a huge temptation to make a player that refuses to play "unauthorized" content, which will wreck the whole scheme. But maybe they are smart enough to resist this.
Besides idiot who will try to make the "won't play unauthorized content player", you mention the real threat to this. The *AA companies are using piracy as an excuse for DRM. They certainly don't want to see piracy stopped as this will remove their argument for DRM. DRM means they can enforce pay-per-view, and eventually (once enough schemes are broken that they can convince the government to make players of non-encrypted content illegal) they can set up a system where "unlicensed" content producers (ie amateurs) cannot distribute entertainment, as distribution requires purchase of an expensive encryption license.
Please everybody, support this as much as possible. This is the real solution to piracy.
Basically, why did they even add the option to their DRM to prevent "squirting"? Did the seriously think that the paranoid recording companies would *not* use that option if it was there? If they had the slightest brains at all, they would have realized that they had in fact disabled one of the few features of the Zune. And they certainly should have known they had the power to refuse to implement such an option. Most of the DRM is there to lock people into using the Zune and thus satisfies Microsoft's own personal interests, but this option makes no sense what so ever!
No matter whether you think Microsoft is Satan's spawn or the greatest company in the world, it is obvious this decision is completely against their desires. This is proof positive that they are a bunch of idiots. This is scary yet relieving at the same time...
Just to be more accurate, the thing that Microsoft really made sure was impossible was dual-boot. They knew that that could lead to alternative os's. At the time I fully expected machines would be sold with Windows plus a "game operating system" and all games would require you to reboot to the game system. Microsoft realized this as well and made absolutely sure that such a machine would never be produced.
Machines that just run an alternative system were never as much of a problem for Microsoft, and today they allow them. They rely on the fact that you need Windows to run Word or Quicken or TurboTax, so any machine that does not run Windows is worthless. So even today they make sure it is impossible to sell a dual-boot machine, while they allow alternative systems so they can pretend they are obeying the monopoly rules.
must be possible to convert existing Word documents to it without any loss. In order to do this, there must be a one-to-one mapping between the .DOC semantics and the OpenXML semantics.
Uh, no, that is wrong. A many-to-one mapping would work for that requirement. If doc has many ways of representing the same information, turning them all into the same way would work just fine and would be lossless.
A requirement that it lossly convert *both* ways, however, would require this. Otherwise converting to XML and back would probably map the many ways of representing the same data to a single one and would be lossy. However I am willing to bet that OOXML does not satisfy that criteria. So this excuse is totally bogus.
Fudge. The second link seems to work. Hit "parent" in it to see the first comment.
The comment is +5:
l y&threshold=1&commentsort=0&mode=thread&pid=176234 74
r eshold=1&commentsort=0&mode=thread&cid=17620482
http://slashdot.org/comments.pl?sid=217018&op=Rep
This response was interesting. I saw it a second time but can't find it now:
http://yro.slashdot.org/comments.pl?sid=217018&th
Take a look above and you will find an equally-irate person complaining that the Democrats got full names and that they thought this was disparaging to the Democrats. This was somewhat backed up by responses which indicated that the original submission had the title "Democrats push new DRM bill" which certainly implies the original wording was done by somebody who did not like Democrats.
The should fix these articles by naming everybody and putting (D) and (R) after each name, so you and that guy will stop throwing hissy fits. Apparently any discrepency in how things are decided is considered an insult by both sides...
Not only that, the chip will also prevent you from singing or humming the music, since that would allow unauthorized listeners to hear it.
Uploaders don't want to hear this, the hardware manufactures don't want to hear it, and Microsoft *certainly* does not want to hear this. But the content guys, if they want their DRM to work, have to stop listening to Microsoft's promises and require the decoding to be done in *hardware*.
A more complex requirement, one that is going to be difficult to explain, is that the hardware API must be completely documented and exposed, so that a Linux driver is easy to create. Obfuscation never works, while exposing everything will mean that any mistakes will be pointed out instantly by hackers trying to get fame by showing how smart they are. Being secretive about the card is *proof* that the design is not sufficently safe and the content industry should not license it. It is also pretty obvious that about 90% of the work in breaking DRM is by people trying to play them on Linux (real pirates (not uploaders) are much more interested in copying the disk image including the DRM, not in decoding it), so this would greatly reduce the number of smart people trying to crack it. Note that a custom driver will be pretty much equivalent to building your own IR remote for a dvd player, it won't do a lot.
My guess is the card will actually be inline between the graphics card and the display. It would replace a keyed area with the video and force the HDCP on and send it out the cable to the display. It probably also needs to directly connect to the disk player so that a disk image could not be fed to it. There may be schemes of key exchange so that such hardware connections are not needed (software only has access to encrypted and un-reusable data). The system api would probably be pretty much the same as the buttons on a remote control.
And the wife of James Baker, a Republican. Or perhaps you forgot? I know how memory is sometimes fragile.
You might want to check your reading comprehension.
He was talking about putting music purchased from a site ONTO a cd. Not taking music from a cd and putting it on the computer/player. Apple allows a standard non-DRM CD to be made from iTunes AAC files.
You want to pretend that Slashdot is closed minded, but it appears you are. Did you actually read the comments, or are you just making assumptions about what they said? I would say you are making assumptions.
I did read the comments and I would say that virtually 100% accuse this guy of spreading FUD and wanting to be able to install spyware. About the only negative comments on MicroSoft were about them trying to lock games into their platform and Vista, which has nothing to do with what the original article was about, he absolutly does not care about the lock in.
I think you are missing the fact that there was *already* a (non GPL) Windows port from TrollTech. This kind of discouraged anybody from porting the X version to Windows. Some people who would want to just bought the Windows version. Others quite likely decided that a fork like that would be a bad idea, since the two Windows versions were unlikely to be compatable. A lot of possible programmers did not like the idea of doing the work to write something that already existed. There may also have been fear that somebody would insert code from the Troll Tech version (you could get the code easily even though it is not GPL) and violate copyrights.
I suspect that if Qt did *not* have a Windows port but just a GPL X version, there would be a GPL Windows version already, probably several years ago.
No, I would pop up a warning that says "SOME FORMATTING MAY BE LOST", just like Microsoft does with EVERY other format it can save to (whether or not anything will be lost). Now maybe, having access to the source code, I would try to be nice and CONVERT (look up the word "convert" if it is foreign to you) to the new format, if it was possible. But that is not important. I CERTAINLY would not put a marker in the program that says "hey, YOU convert for me, as I don't want to!".
What they are doing is indefensible, so don't even try.
The problem with that solution is that it will allow software other than MSWord to read the saved documents correctly by making the spec simple enough to implement. Though not in the ISO submitted documentation, "make sure it is impossible for other software to read this correctly" was one of the primary design criteria, so the solution you propose was eliminated immediately.
No, MSWord should pop up a scary-sounding warning like "SOME FORMATTING MAY BE LOST" and save the file the best it can.
Rest assurred that if they implemented ODF it would pop up the most scary warnings possible, saying the user's very life may be threatened if they save this way, even if nothing would be lost. They should apply the same rules to their own code, and stop this bullshit.
No to both.
Searching UTF-8 works by using string subsearch, which has the advantage that it can also search for any word, not just some glyph. Also you should check real code, I very much doubt anybody searches for any single unicode character other than some ASCII punctuation, which could be searched for in UTF-8 by the normal ASCII search (since it won't match any bytes of a multibyte character).
UCS-4 (the correct name for UTF-32 because no actual transformation is done) is useful, but mostly as for graphics when actually drawing the glyphs. I recommend that all storage be done in UTF-8, mostly because it is extremely difficult to eliminate all ASCII internally, so this is the only way to make all your string data be the same.
The glib people apparently though UCS-4 would be useful, that is why wchar_t on Linux is 4 bytes, instead of the 2 used by everybody else, including other Unix implementations. However the main effect is that this means code using wchar_t is incompatable between Windows and Linux, and that nobody on Linux uses it. About the only positive effect I see is that it discourages wchar_t usage.
I believe Unicode does have problems, but "surrogate pairs" is strictly a problem with the UTF-16 encoding chosen by Microsoft (and by the Unix companies before that) because they failed to realize the impossibility of having fixed-size characters. Plan9 chose UTF-8. Linux pretty much chose it as well, I would like to think it is because they were smart, but the main reason is that it is enoumously easier to implement UTF-8 on an existing system than to add "wide characters".
Unicode's problems are chiefly the insistence that all existing character sets have a lossless translation from/to Unicode. This is why there are hundreds of combined character+accent mark glyphs, when it would be a lot easier for everybody if only decomposed characters were supported. I think there may be similar problems with all the pictrograph characters for Chinese and Japanese, in that they could instead be "composed" of several "stroke" characters, but I certainly can't say anything about languages I know nothing about.
A good deal of problems with Unicode and encodings I think stem from a politically-correct view that it is somehow bad if the Americans get the "better" short characters. Thus the insistence that things all be equal-sized when stored and that characters be composed.
What language is the VM written in?
You are apparently totally ignorant of modern design if you think storing strings as bytes preculdes Unicode. There is something called UTF-8. Look it up sometimes. You might need to hit yourself with the cluestick a few times to remove the delusion that there is some advantage in having all the glyphs take the same number of bits and/or that there is any solution out there where the glyphs do all take the same number of bits. Hints: 1.Text is made of WORDS, sentences, paragraphs, lines, and many many other structures that are variable length. 2.There are things called "surrogate pairs" in utf-16. 3.There are "combining characters".
The fact that D jettisoned a whole lot of crap for "wide characters" is one of the best indications that they get it and know what they are doing!
The problem is not coming up with a question that a computer program cannot answer. The problem is making a computer program that can create such questions.
If the questions are created by a human, there is going to be a limited set. The spammers only have to figure out the answers to that limited set. Only by having the computer generate a (essentially) infinite set of questions can this workaround be avoided.
Anybody who trusts security through obscurity is an idiot, and you have just shown you are one.
Any proper design MUST assume the bad guys have every single bit of information about the machine. Open source software may be a way to make sure the designers do not miss this assumption, and thus is extremely good for security of the machine.
No, that is a lie that is actually being propagated by conservatives (!) to stop investigation of electronic voting, by making people think it will upset the entire country and threaten the conservative leadership.
In fact the last election would be an *excellent* time to investigate the electronic voting, for precisely the reason that it *can't* affect the outcome. Even if you assumme every single electronic machine in Ohio was fixed to throw the election to the Republicans, it would require the counties with those machines to really be voting something like 70% for Kerry, while adjacent and otherwise identical counties without electronic machines somehow were voting 49% for Kerry, for the election to come out differently.
Of course the conservatives are scared to death that something funny did happen and will be detected, so they will make sure any attempt to detect it is avoided by a bogus threat that it will destabilize the country. And a bunch of nutjob Liberals who refuse to admit that perhaps the majority in the country don't agree with them are doing there best to help this plot along...
BeOS was of course killed by Microsoft refusing to allow OEM's to sell dual-boot machines.
I'm sure lots of other stuff was killed without even seeing the light of day because of this. Just before BeOS, I fully expected machines that would run Windows and that you rebooted in order to play games into a special game OS, which would allow another company making that system to compete, and eventually the Game OS would become so powerful that Microsoft would have real competetion. But Microsoft stopped that dead by not allowing dual-boot. The only reason you heard of BeOS is that they thought their system was so good that maybe they could also sell non-dual-boot machines.
Plenty of people have asked where the preinstalled Linux machines are. Well they somewhat exist, businesses buy them. But just check what *everybody* using desktop Linux really has, and you will find that 100% of them are dual-boot machines. Microsoft is not stupid, they know that dual-boot is the real killer and will do anything they can to stop it.