If you want to claim copyright for your own works, you have to honor the copyright claims of others
In order for that sentence to make any sense at all, two things must be true: copyright must exist, and 'you' have a CHOICE (notice the IF in your statement) as to whether you want enforce it or not.
Copyright as a concept must exist, but that goes without saying. That does not imply that it represents any kind of universal right which you could enforce on others with threats of fines or other violation of their physical property rights or personal liberty. As always, reciprocation is the key: it would be nonsensical to claim copyright protection on your own works while ignoring similar claims made by others, and the same goes for their claims if they choose to ignore yours.
if you do not attempt to claim copyright on your own works, others cannot press any copyright claims against you
So, copyright still exists, you still have a choice as to whether or not to enforce, but NOBODY ELSE has the same choice.
On the contrary, everyone else has exactly the same choice: respect others' copyright claims and have a reasonable expectation that they (meaning the others making copyright claims) will respect yours, or ignore others' copyright claims and have your own (if any) ignored in turn. This is how rights work: you respect others' and they respect yours in turn, because doing otherwise would undermine both your claims. As a consequence, rights are only meaningful so long as they are valued by both sides. This is what prevents them from being completely arbitrary.
Copyright IS uniform and universal.... EVERYBODY has the right to create things, and EVERYBODY gets the SAME protections for the things they created.
Protections which are valued only by the pro-copyright faction, and thus not universal. If you are for copyright you receive the protection you wanted. If you are against copyright and do not wish these protections for your own work then you receive nothing of value, rather the contrary: those protections have negative value to you.
It is not enough for a right to be uniformly applied; it also has to be equitable. The cost of respecting the right as it applies to others should be perceived by all involved as a fair trade in exchange for others' respecting the right as it applies to oneself. Rights are a voluntary arrangement, not objective fact: anyone can choose to opt out of any right at any time, simply by choosing not to respect the right as it applies to others. The rights which are considered universal, like physical property and personal liberty, are the ones that few would choose to opt out of because everyone depends on them to function. Copyright, on the other hand, is not such a right. Opting out would involve little in the way of consequence for the majority of the population, whether because they produce nothing of value, or (more commonly) because they do not seek to restrict others from benefiting from or distributing the creative works which they do produce.
So you think you can take away my rights by not using yours?
You mischaracterize the argument by phrasing the rejection of copyright as "not using" copyright. This is a form of circular reasoning; you presume the validity of copyright law from the start, when this is the very issue at stake. You claim that the right exists; others disagree. You are naturally free to behave as though the right exists when interacting with other like-minded individuals, but there is nothing inherent to the act of copyright violation that would justify a punishment more severe than violating the other party's copyrights in turn, which is of course no punishment at all for anyone outside your pro-copyright group who claims no copyrights in the first place.
The only legitimate rights are the ones which can be applied uniformly and universally, and only justifiable responses to the violation of a right are the ones proportional to the violation—meaning similar in both form and magnitude. Fines and/or jail time involve the violation of rights which rest on a completely different philosophical foundation than copyright and would consequently be disproportionate to the alleged offense.
A fine solution, if it were actually implemented consistently. If you want to claim copyright for your own works, you have to honor the copyright claims of others. Reciprocation is fair, after all. However, if you do not attempt to claim copyright on your own works, others cannot press any copyright claims against you. The penalty for copyright infringement is thus limited to having your own copyright claims invalidated.
The problem with copyright law is that it affects everyone, whether they want it or not. Asymmetric law is inherently unjust.
Treaties do NOT supersede the constitution. Where the hell did you get that idea from?
This Constitution, and the Laws of the United States which shall be made in Pursuance thereof; and all Treaties made, or which shall be made, under the Authority of the United States, shall be the supreme Law of the Land
Sounds like they got the idea from Article 6.
Only if "they" lack reading comprehension. The quoted passage does not say that treaties supersede the Constitution, only that treaties make up one part of the supreme law of the land, alongside the Constitution and U.S. law. Note in particular that U.S. law is given the same status, and if U.S. law superseded the Constitution then there would be no need for a constitutional amendment process, as Congress would only need to pass a law to override any part of the Constitution.
No, object code is not source code even if you write it by hand.
Whether it is technically "source code" is debatable, but not really relevant. The point is that it is in the original (source) language and does not require translation to any other language.
Implementing your translator in hardware doesn't mean it's not a translator.
Right, but you don't need to implement a translator at all, hardware or software. Interpretation (which is what a CPU does with ordinary object code) is not translation. Any programming language can be executed directly by hardware without first being translated to a separate object code language; it just requires a much more complicated hardware design, particularly in the instruction decode unit, which is why programs are generally compiled to a simpler language. Von Neumann architecture (or the more modern Harvard architecture) would work just fine. I'm not sure why you think a neural net would be useful, much less necessary, when ordinary compilers manage quite well without them.
The boundary between hardware and software is a flexible one
Not really. Hardware is gates and memory cells.
True, but irrelevant. Any software program can be synthesized in the form of gates and memory cells, thus becoming hardware, and any circuit composed of gates and memory cells can be modeled and emulated in software.
Tell me, for a program with an arbitrary number of variables and dynamic memory allocation, how many registers do you think you'd need?
None. The program can keep its data directly in RAM, with variable names taking the place of virtual addresses. Some architectures (e.g. PowerPC) only allow computation via registers, but others (like x86) are able to use memory locations as operands directly. The size of the register file is just another level of optimization.
All of this would no doubt be horribly inefficient; speaking as someone who has done plenty of low-level software (including some hand-written object code) as well as hardware (basic register- and stack-oriented CPUs and peripherals synthesized for FPGAs) and a fair amount of parsing and program translation, there are perfectly good reasons why computers use distinct object code languages and compilers rather than interpreting high-level code directly. However, this is a matter of design choices and not any substantial qualitative difference between source code and object code—circuits of "gates and memory cells" could, in principle, be designed to interpret either form directly.
Well, object code has to be directly executable by a processing architecture.
Putting aside the fact that the source code and object code may be identical—as humans can, and sometimes do, write object code directly—there is no reason that high-level source code, e.g C# or JavaScript, could not be executed directly by the hardware with no intermediate translation step. The boundary between hardware and software is a flexible one; anything you can do in software (like interpreting source code) can also be realized with special-purpose hardware, just as the vast majority of opcodes in a modern CPU could be implemented in software on top of a far smaller set of primitives (e.g. a Turing machine). The current split between hardware and software (and compilation vs. interpretation) is merely the result of an optimization process, applying each style to the area(s) it is best suited for.
I'm not altogether I'd call source code "speech", but I'm pretty sure that object code isn't speech. If Apple wanted to use a binary editor to change the number of tries in the object code from 10 to 10,000, then sign the result, they haven't been forced to utter speech.
There is no fundamental difference between source code and object code, aside from the level of abstraction. Both are lists of instructions and associated data. Both can be read and understood by both computers and (properly trained) humans.
The signature, however, is the bigger issue. Forcing Apple to sign a compromised build of the software with their code-signing key is exactly equivalent to forcing them to publicly attest that the compromised build is approved by Apple for operation on the iPhone 5C. That is what it means for the software to be signed with their key. It isn't just a technical measure, it's a combination opinion statement and promise from Apple that the build is suitable for use on the iPhone.
Aside from all of that, this court order is effectively a "writ of assistance", and the use (and abuse) of such writs to compel outside parties to assist the courts and law enforcement in their operations was one of the complaints which led the colonies to rebel against Britain and found the United States in the first place. This is not a practice we can afford to reintroduce.
Gold is great until you need more of it, which you always do, because economies grow (all being well) and extra money is needed to support that.
You do need the total supply of gold to become more valuable as the economy grows. What you do not need, however, is more gold. Absent deliberate interference, the law of supply and demand ensures that the value of the existing gold adjusts automatically to match the value of everything it can be used to purchase.
The only reason to need more currency would be if the supply were stretched so thin that the units required for common transactions became impractically small. However, in cases like that you can just substitute a more common material (like silver) for everyday use and reserve the gold for large transactions. However, gold's physical properties make it fairly easy to handle even in small quantities; for example, you can purchase 1/20 of a gram (about $4 worth) of gold wire embedded in a plastic card from Shire Silver, which is no harder to store or transact with than a typical $5 paper note.
Bitcoin, of course, does not suffer from this problem at all since it is just as easy to transfer 0.00000001 XBT (one satoshi) as it is to transfer 1 XBT—and if that ever did become an issue it would not be difficult to add a few extra decimal places to the protocol.
I thought I read somewhere that this device has a JTAG connector somewhere inside it. Seems reasonable to me that they could read out the memory content with that...
You seem to think that, contrary to all evidence, the designers of the Secure Enclave chip must have been really stupid if you think they would even consider allowing JTAG access to the memory holding the device-specific key which is never, under any circumstances, to leave the Secure Enclave, thus nullifying the entire purpose of the chip.
I don't have all the details but I do think that someone who commits a crime and completes the punishment should be able to get back to a somewhat normal life.
So you believe that the fact that this person did commit the crime is no longer relevant. That is a perfectly reasonable and even admirable position to take. It does not, however, justify censoring information about his past. This is an attempt to apply an unjust legal solution to what is fundamentally a social problem. The solution to social problems is persuasion, not censorship. Others have the right to make their own decisions based on all the available data. Rather than actively interfering with that process, you need to convince them that the person's past is truly irrelevant.
Personally, I find it doubtful that anyone who was truly rehabilitated would choose to hide their past crimes from others. Part of rehabilitation involves accepting your past and the consequences it holds for your relationships with others, not covering it up and living a lie in constant fear of discovery.
While I agree with you that all businesses should have the right to refuse service, regardless of the reason, those cases really aren't the same as the situation of being compelled to produce a product with a message that one disagrees with. In the three cases you cited, the businesses refused to provide the same services to same-sex couples which they already provide to everyone else. There was no message involved, unless you count the implied endorsement of same-sex marriage which may or may not attach to simply providing a generic wedding cake or flowers or a place to hold the ceremony.
It is unfortunate that one apparently cannot provide a useful and beneficial service to most, but not all, members of the public without having one's labor and capital effectively appropriated by the State for the public use of the remainder. If you do not wish to be compelled to facilitate activities you personally find abhorrent, the only option the State offers you is to avoid providing the service at all, to society's detriment.
Justice Stevens' dissenting opinion on the 1977 United States, Petitioner, v. New York Telephone Company case is as insightful and applicable to smartphone backdoors today as it was in 1977 when the court ordered a telephone company to provide technical assistance with installing a pen register. In particular:
If the All Writs Act confers authority to order persons to aid the Government in the performance of its duties, and is no longer to be confined to orders which must be entered to enable the court to carry out its functions, it provides a sweeping grant of authority entirely without precedent in our Nation's history. Of course, there is precedent for such authority in the common law the writ of assistance. The use of that writ by the judges appointed by King George III was one British practice that the Revolution was specifically intended to terminate.... I can understand why the Court today does not seek to support its holding by reference to that writ, but I cannot understand its disregard of the statutory requirement that the writ be "agreeable to the usages and principles of law."
(The rest of the dissent is also well worth reading.) Even without an actual repeal of the All Writs Act, the Act should never have been interpreted so broadly as to cover what the court is demanding from Apple.
I believe the requirement of a public Writ signed off by a judge is what is making it rule of law with checks and balances, and not a police state.
You don't think that police states (some of them, anyway) include courts willing to sign off on whatever the police want to do? Judicial involvement and possession of a perfectly legal public Writ does not in any way detract from the fact that the "necessities" of law enforcement are prioritized over individual rights—the essence of a police state.
The "zombie horde" is going to make the choice for you either way. Joining them in the process does nothing but add your own stamp of legitimacy to the outcome they picked.
If you believe that others have the right to interfere in your life just because they have an advantage in numbers, you have no basis to complain about the result. If not, how can you justify voting yourself? Either you lose, in which case you accomplish nothing, or you win and your chosen candidate interferes in others' lives in exactly the same way you despise when it's happening to you.
If the tenant receives a court order to open the lock, he will do.
Only to avoid the expense of replacing the door. The warrant gives the police the authority to break in; it's up to them to find the means to do so. The warrant imposes no obligation on the owner, much less any third party (like Apple), beyond staying out of the officers' way while they conduct their search.
I'd assume that for any random Android device it's possible for someone (oem, carrier, Google) to unlock the bootloader, allowing a custom recovery which can, basically, do what they've ordered Apple to allow with the iPhone.
Sure, you can unlock the bootloader; just connect to a PC and run "fastboot oem unlock". However, both locking and unlocking the bootloader have the effect of a factory reset, so this won't allow you to get at the user's data.
as somebody figures out your wallet id which you need to share with anybody you transfer a coin from or to, they can go get your entire history
That's not how Bitcoin works. You can have any number of wallets, and most Bitcoin clients support the new "HD" (Hierarchical Deterministic) wallet format, where many "IDs" are generated from a master key using a one-way function. A single key can provide an infinite stream of distinct addresses, one for each transaction. Without the master key there is nothing to link these addresses together. You don't give people "your wallet ID", you give them an address unique to that one transaction. When you spend the bitcoins later you send the change to a different unique address; if you do it right then it won't be obvious to third-parties which output was the expenditure and which was the change.
The one case that is more difficult to avoid is when you need to combine multiple inputs together to cover a single larger expenditure. The fact that the inputs were used in the same transaction is fair evidence that they are controlled by the same entity. This is where schemes like CoinJoin and mixing services can help to preserve anonymity, by combining many unrelated inputs and redistributing the sums to new addresses in standardized denominations, thus breaking the link between the inputs and the outputs. It is possible to implement this using blinded signatures such that even the mixing service doesn't which inputs are linked to which outputs.
Then there is the fact that "your entire history" is limited to your transactions on the blockchain, which are still pseudonymous for both senders and receivers—it would take a whole lot of old-fashioned detective work to determine the real-world identities of everyone you've traded with even after their (one-time-use) Bitcoin addresses are known.
Have you considered the possibility that for many people that may be a very good trade? Suppose the vast majority of citizens, let's say 70% for example, are quite OK with trading liberties for security. So what exactly do you do then?
Even if there was only one person in the entire world who preferred liberty, that would still be their right, no matter how many others are perfectly willing to trade it away for the promise of security. You have the right to forfeit your own liberty, if that is your choice; you do not have the right to trade away others' liberty, no matter how outnumbered they may be.
The purpose of a progressive tax code is redistribution, not fairness.
There is nothing inherent in progressive tax code that leads to redistribution. That is on the spending side of government, not the taxation side.
Redistribution depends on both taxation and spending, specifically whether the difference between the amount you pay in taxes and the amount you benefit from the spending is correlated with wealth and/or income. I had already argued that the increased taxes under a progressive tax policy are not balanced by an increase in benefits, which leaves as the only other option a situation where the difference is redistributed to others who pay less in taxes and yet benefit more from the spending.
As for insurance, if you buy a car that cost three times as much as I do, you would expect to pay three times the insurance, assume the other factors being the same. And yet, with taxes, at least in the US, it does not work that way.
The "taxes as insurance" analogy isn't a very good one—unless you mean "insurance" as in "protection racket", then it's fairly apt—but also consider this: if you can afford to lose the car, you don't need (comprehensive) insurance on it at all. You'll still need liability insurance, of course, but that's for a fixed amount and doesn't depend on the value of the vehicle (and if you're sufficiently wealthy you may be able to self-insure for liability as well).
If taxes were analogous to insurance then the rates would be regressive, not progressive. The wealthier you are the less protection you need, proportionally speaking. You have more to lose, but you can afford to lose more without serious hardship. The wealthy do not require proportionally greater levels of government services than the poor, and can easily afford to provide those services for themselves. The purpose of a progressive tax code is redistribution, not fairness.
I find it doubtful this is technically impossible as Apple are making out.... This capability certainly exists in their test labs otherwise every update would be at risk of bricking phone on general release.
The risk isn't "bricking" the phone, just erasing the storage and starting over. Not a big deal, especially in a test lab environment.
This particular phone may be just barely old enough that the security features are implemented in software. Newer models, starting with the iPhone 5S, have already fixed that flaw by moving features like the minimum delay between attempts and auto-erasing the key after 10 failures into the Secure Enclave. Also, newer versions do not permit updates to be installed while the device is locked. If the phone were just one model newer then it would be demonstrably impossible for Apple to produce a software update to circumvent the security. As it is, they might be able to comply—but they shouldn't.
What this case is about what a third party can reasonably be ordered to do (without compensation?) to facilitate a legal search. This goes well beyond a landlord being ordered to unlock a back door. Or even allowing a wire tap to be installed on a phone line. My guess would be that assisting the FBI would probably take a few days and potentially disrupt Apple's iOS QA cycle for that long if they have to utilize in house resources.
Maybe longer since they essentially have to fork the iOS code base for this one device and then somehow isolate and target this one device for a software update. Oh and really trying hard not to brick the phone in the process. Not trivial, but certainly a somewhat borderline case considering the relatively vast resources of Apple.
All true, but I would argue that the fact that they're being asked to do something which they find morally repugnant, and the potential damage to their reputation, are greater considerations than the marginal costs in time, labor, and materials, which they would likely be (partly) compensated for in any event. Signing a firmware build is not just a technical measure to carry out a warrant; it's a statement by Apple that this build of the firmware has their seal of approval and is authorized to run on the iPhone. They have not even been accused of breaking any law, much less doing anything actually immoral, and yet they are being ordered by the court to provide false testimony under duress—all for the sake of gaining access to a dead suspect's work phone which probably doesn't contain any information pertinent to the investigation.
Well, that probably beats ROT-13, though I wouldn't care to stake anything important on it.
However, it is most definitely not a one-time pad. Despite being provably unbreakable when implemented correctly, OTP is perhaps the simplest of all encryption algorithms. There is no hashing, key-stretching, chaining, etc., just a message and a true-random pad of the same length. To encrypt or decrypt you just XOR the message and the pad.
The difficulty, of course, is in the key exchange. You have to arrange for both parties to securely share the same random pad, and any given bit of the pad can only be used to communicate one bit of message. Moreover, every pair of endpoints requires a different pad—there is no OTP equivalent of public-key cryptography.
This is where quantum cryptography really shines, as it employs quantum effects to negotiate random pads in real time between two endpoints without risk of undetected eavesdropping. The encryption algorithm in quantum cryptography is really just OTP; the novel part is the solution to the key exchange problem.
Actually, there is a hack for that. I saw a demonstration of a device that uses the USB interface to try a small number of PINs, then resets the device before it records the failures, then tries another small number of PINs, and so on.
This is why you should record the attempt first, before checking whether the PIN is valid. Resetting the device at any point in the authentication process before the attempt has been recorded should not yield any additional information about the PIN.
In this case, we have a lawful order from a court. Legally and morally, Apple should comply.
Legally, perhaps. The order may well be unlawful; compelling a third party not under investigation to produce brand-new software to circumvent existing security measures and sign it with their seal of approval as an approved software update is well outside the bounds of what these order are normally used for.
Morally, even if the order turns out to be legal, Apple should refuse. Regardless of its legality, the order itself is immoral and should not be followed. Following the order would be harmful to the security of every iPhone 5c owner, the vast majority of whom are innocent bystanders and not accused of any wrongdoing.
Slashdot Mirror
If you want to claim copyright for your own works, you have to honor the copyright claims of others
In order for that sentence to make any sense at all, two things must be true: copyright must exist, and 'you' have a CHOICE (notice the IF in your statement) as to whether you want enforce it or not.
Copyright as a concept must exist, but that goes without saying. That does not imply that it represents any kind of universal right which you could enforce on others with threats of fines or other violation of their physical property rights or personal liberty. As always, reciprocation is the key: it would be nonsensical to claim copyright protection on your own works while ignoring similar claims made by others, and the same goes for their claims if they choose to ignore yours.
if you do not attempt to claim copyright on your own works, others cannot press any copyright claims against you
So, copyright still exists, you still have a choice as to whether or not to enforce, but NOBODY ELSE has the same choice.
On the contrary, everyone else has exactly the same choice: respect others' copyright claims and have a reasonable expectation that they (meaning the others making copyright claims) will respect yours, or ignore others' copyright claims and have your own (if any) ignored in turn. This is how rights work: you respect others' and they respect yours in turn, because doing otherwise would undermine both your claims. As a consequence, rights are only meaningful so long as they are valued by both sides. This is what prevents them from being completely arbitrary.
Copyright IS uniform and universal. ... EVERYBODY has the right to create things, and EVERYBODY gets the SAME protections for the things they created.
Protections which are valued only by the pro-copyright faction, and thus not universal. If you are for copyright you receive the protection you wanted. If you are against copyright and do not wish these protections for your own work then you receive nothing of value, rather the contrary: those protections have negative value to you.
It is not enough for a right to be uniformly applied; it also has to be equitable. The cost of respecting the right as it applies to others should be perceived by all involved as a fair trade in exchange for others' respecting the right as it applies to oneself. Rights are a voluntary arrangement, not objective fact: anyone can choose to opt out of any right at any time, simply by choosing not to respect the right as it applies to others. The rights which are considered universal, like physical property and personal liberty, are the ones that few would choose to opt out of because everyone depends on them to function. Copyright, on the other hand, is not such a right. Opting out would involve little in the way of consequence for the majority of the population, whether because they produce nothing of value, or (more commonly) because they do not seek to restrict others from benefiting from or distributing the creative works which they do produce.
So you think you can take away my rights by not using yours?
You mischaracterize the argument by phrasing the rejection of copyright as "not using" copyright. This is a form of circular reasoning; you presume the validity of copyright law from the start, when this is the very issue at stake. You claim that the right exists; others disagree. You are naturally free to behave as though the right exists when interacting with other like-minded individuals, but there is nothing inherent to the act of copyright violation that would justify a punishment more severe than violating the other party's copyrights in turn, which is of course no punishment at all for anyone outside your pro-copyright group who claims no copyrights in the first place.
The only legitimate rights are the ones which can be applied uniformly and universally, and only justifiable responses to the violation of a right are the ones proportional to the violation—meaning similar in both form and magnitude. Fines and/or jail time involve the violation of rights which rest on a completely different philosophical foundation than copyright and would consequently be disproportionate to the alleged offense.
Let copyright exist for those who want it.
A fine solution, if it were actually implemented consistently. If you want to claim copyright for your own works, you have to honor the copyright claims of others. Reciprocation is fair, after all. However, if you do not attempt to claim copyright on your own works, others cannot press any copyright claims against you. The penalty for copyright infringement is thus limited to having your own copyright claims invalidated.
The problem with copyright law is that it affects everyone, whether they want it or not. Asymmetric law is inherently unjust.
Treaties do NOT supersede the constitution. Where the hell did you get that idea from?
This Constitution, and the Laws of the United States which shall be made in Pursuance thereof; and all Treaties made, or which shall be made, under the Authority of the United States, shall be the supreme Law of the Land
Sounds like they got the idea from Article 6.
Only if "they" lack reading comprehension. The quoted passage does not say that treaties supersede the Constitution, only that treaties make up one part of the supreme law of the land, alongside the Constitution and U.S. law. Note in particular that U.S. law is given the same status, and if U.S. law superseded the Constitution then there would be no need for a constitutional amendment process, as Congress would only need to pass a law to override any part of the Constitution.
No, object code is not source code even if you write it by hand.
Whether it is technically "source code" is debatable, but not really relevant. The point is that it is in the original (source) language and does not require translation to any other language.
Implementing your translator in hardware doesn't mean it's not a translator.
Right, but you don't need to implement a translator at all, hardware or software. Interpretation (which is what a CPU does with ordinary object code) is not translation. Any programming language can be executed directly by hardware without first being translated to a separate object code language; it just requires a much more complicated hardware design, particularly in the instruction decode unit, which is why programs are generally compiled to a simpler language. Von Neumann architecture (or the more modern Harvard architecture) would work just fine. I'm not sure why you think a neural net would be useful, much less necessary, when ordinary compilers manage quite well without them.
The boundary between hardware and software is a flexible one
Not really. Hardware is gates and memory cells.
True, but irrelevant. Any software program can be synthesized in the form of gates and memory cells, thus becoming hardware, and any circuit composed of gates and memory cells can be modeled and emulated in software.
Tell me, for a program with an arbitrary number of variables and dynamic memory allocation, how many registers do you think you'd need?
None. The program can keep its data directly in RAM, with variable names taking the place of virtual addresses. Some architectures (e.g. PowerPC) only allow computation via registers, but others (like x86) are able to use memory locations as operands directly. The size of the register file is just another level of optimization.
All of this would no doubt be horribly inefficient; speaking as someone who has done plenty of low-level software (including some hand-written object code) as well as hardware (basic register- and stack-oriented CPUs and peripherals synthesized for FPGAs) and a fair amount of parsing and program translation, there are perfectly good reasons why computers use distinct object code languages and compilers rather than interpreting high-level code directly. However, this is a matter of design choices and not any substantial qualitative difference between source code and object code—circuits of "gates and memory cells" could, in principle, be designed to interpret either form directly.
Well, object code has to be directly executable by a processing architecture.
Putting aside the fact that the source code and object code may be identical—as humans can, and sometimes do, write object code directly—there is no reason that high-level source code, e.g C# or JavaScript, could not be executed directly by the hardware with no intermediate translation step. The boundary between hardware and software is a flexible one; anything you can do in software (like interpreting source code) can also be realized with special-purpose hardware, just as the vast majority of opcodes in a modern CPU could be implemented in software on top of a far smaller set of primitives (e.g. a Turing machine). The current split between hardware and software (and compilation vs. interpretation) is merely the result of an optimization process, applying each style to the area(s) it is best suited for.
I'm not altogether I'd call source code "speech", but I'm pretty sure that object code isn't speech. If Apple wanted to use a binary editor to change the number of tries in the object code from 10 to 10,000, then sign the result, they haven't been forced to utter speech.
There is no fundamental difference between source code and object code, aside from the level of abstraction. Both are lists of instructions and associated data. Both can be read and understood by both computers and (properly trained) humans.
The signature, however, is the bigger issue. Forcing Apple to sign a compromised build of the software with their code-signing key is exactly equivalent to forcing them to publicly attest that the compromised build is approved by Apple for operation on the iPhone 5C. That is what it means for the software to be signed with their key. It isn't just a technical measure, it's a combination opinion statement and promise from Apple that the build is suitable for use on the iPhone.
Aside from all of that, this court order is effectively a "writ of assistance", and the use (and abuse) of such writs to compel outside parties to assist the courts and law enforcement in their operations was one of the complaints which led the colonies to rebel against Britain and found the United States in the first place. This is not a practice we can afford to reintroduce.
Gold is great until you need more of it, which you always do, because economies grow (all being well) and extra money is needed to support that.
You do need the total supply of gold to become more valuable as the economy grows. What you do not need, however, is more gold. Absent deliberate interference, the law of supply and demand ensures that the value of the existing gold adjusts automatically to match the value of everything it can be used to purchase.
The only reason to need more currency would be if the supply were stretched so thin that the units required for common transactions became impractically small. However, in cases like that you can just substitute a more common material (like silver) for everyday use and reserve the gold for large transactions. However, gold's physical properties make it fairly easy to handle even in small quantities; for example, you can purchase 1/20 of a gram (about $4 worth) of gold wire embedded in a plastic card from Shire Silver, which is no harder to store or transact with than a typical $5 paper note.
Bitcoin, of course, does not suffer from this problem at all since it is just as easy to transfer 0.00000001 XBT (one satoshi) as it is to transfer 1 XBT—and if that ever did become an issue it would not be difficult to add a few extra decimal places to the protocol.
I thought I read somewhere that this device has a JTAG connector somewhere inside it. Seems reasonable to me that they could read out the memory content with that...
You seem to think that, contrary to all evidence, the designers of the Secure Enclave chip must have been really stupid if you think they would even consider allowing JTAG access to the memory holding the device-specific key which is never, under any circumstances, to leave the Secure Enclave, thus nullifying the entire purpose of the chip.
I don't have all the details but I do think that someone who commits a crime and completes the punishment should be able to get back to a somewhat normal life.
So you believe that the fact that this person did commit the crime is no longer relevant. That is a perfectly reasonable and even admirable position to take. It does not, however, justify censoring information about his past. This is an attempt to apply an unjust legal solution to what is fundamentally a social problem. The solution to social problems is persuasion, not censorship. Others have the right to make their own decisions based on all the available data. Rather than actively interfering with that process, you need to convince them that the person's past is truly irrelevant.
Personally, I find it doubtful that anyone who was truly rehabilitated would choose to hide their past crimes from others. Part of rehabilitation involves accepting your past and the consequences it holds for your relationships with others, not covering it up and living a lie in constant fear of discovery.
While I agree with you that all businesses should have the right to refuse service, regardless of the reason, those cases really aren't the same as the situation of being compelled to produce a product with a message that one disagrees with. In the three cases you cited, the businesses refused to provide the same services to same-sex couples which they already provide to everyone else. There was no message involved, unless you count the implied endorsement of same-sex marriage which may or may not attach to simply providing a generic wedding cake or flowers or a place to hold the ceremony.
It is unfortunate that one apparently cannot provide a useful and beneficial service to most, but not all, members of the public without having one's labor and capital effectively appropriated by the State for the public use of the remainder. If you do not wish to be compelled to facilitate activities you personally find abhorrent, the only option the State offers you is to avoid providing the service at all, to society's detriment.
Justice Stevens' dissenting opinion on the 1977 United States, Petitioner, v. New York Telephone Company case is as insightful and applicable to smartphone backdoors today as it was in 1977 when the court ordered a telephone company to provide technical assistance with installing a pen register. In particular:
If the All Writs Act confers authority to order persons to aid the Government in the performance of its duties, and is no longer to be confined to orders which must be entered to enable the court to carry out its functions, it provides a sweeping grant of authority entirely without precedent in our Nation's history. Of course, there is precedent for such authority in the common law the writ of assistance. The use of that writ by the judges appointed by King George III was one British practice that the Revolution was specifically intended to terminate. ... I can understand why the Court today does not seek to support its holding by reference to that writ, but I cannot understand its disregard of the statutory requirement that the writ be "agreeable to the usages and principles of law."
(The rest of the dissent is also well worth reading.) Even without an actual repeal of the All Writs Act, the Act should never have been interpreted so broadly as to cover what the court is demanding from Apple.
I believe the requirement of a public Writ signed off by a judge is what is making it rule of law with checks and balances, and not a police state.
You don't think that police states (some of them, anyway) include courts willing to sign off on whatever the police want to do? Judicial involvement and possession of a perfectly legal public Writ does not in any way detract from the fact that the "necessities" of law enforcement are prioritized over individual rights—the essence of a police state.
The "zombie horde" is going to make the choice for you either way. Joining them in the process does nothing but add your own stamp of legitimacy to the outcome they picked.
If you believe that others have the right to interfere in your life just because they have an advantage in numbers, you have no basis to complain about the result. If not, how can you justify voting yourself? Either you lose, in which case you accomplish nothing, or you win and your chosen candidate interferes in others' lives in exactly the same way you despise when it's happening to you.
How is it different from a lock on a door?
It isn't.
If the tenant receives a court order to open the lock, he will do.
Only to avoid the expense of replacing the door. The warrant gives the police the authority to break in; it's up to them to find the means to do so. The warrant imposes no obligation on the owner, much less any third party (like Apple), beyond staying out of the officers' way while they conduct their search.
I'd assume that for any random Android device it's possible for someone (oem, carrier, Google) to unlock the bootloader, allowing a custom recovery which can, basically, do what they've ordered Apple to allow with the iPhone.
Sure, you can unlock the bootloader; just connect to a PC and run "fastboot oem unlock". However, both locking and unlocking the bootloader have the effect of a factory reset, so this won't allow you to get at the user's data.
as somebody figures out your wallet id which you need to share with anybody you transfer a coin from or to, they can go get your entire history
That's not how Bitcoin works. You can have any number of wallets, and most Bitcoin clients support the new "HD" (Hierarchical Deterministic) wallet format, where many "IDs" are generated from a master key using a one-way function. A single key can provide an infinite stream of distinct addresses, one for each transaction. Without the master key there is nothing to link these addresses together. You don't give people "your wallet ID", you give them an address unique to that one transaction. When you spend the bitcoins later you send the change to a different unique address; if you do it right then it won't be obvious to third-parties which output was the expenditure and which was the change.
The one case that is more difficult to avoid is when you need to combine multiple inputs together to cover a single larger expenditure. The fact that the inputs were used in the same transaction is fair evidence that they are controlled by the same entity. This is where schemes like CoinJoin and mixing services can help to preserve anonymity, by combining many unrelated inputs and redistributing the sums to new addresses in standardized denominations, thus breaking the link between the inputs and the outputs. It is possible to implement this using blinded signatures such that even the mixing service doesn't which inputs are linked to which outputs.
Then there is the fact that "your entire history" is limited to your transactions on the blockchain, which are still pseudonymous for both senders and receivers—it would take a whole lot of old-fashioned detective work to determine the real-world identities of everyone you've traded with even after their (one-time-use) Bitcoin addresses are known.
Have you considered the possibility that for many people that may be a very good trade? Suppose the vast majority of citizens, let's say 70% for example, are quite OK with trading liberties for security. So what exactly do you do then?
Even if there was only one person in the entire world who preferred liberty, that would still be their right, no matter how many others are perfectly willing to trade it away for the promise of security. You have the right to forfeit your own liberty, if that is your choice; you do not have the right to trade away others' liberty, no matter how outnumbered they may be.
The purpose of a progressive tax code is redistribution, not fairness.
There is nothing inherent in progressive tax code that leads to redistribution. That is on the spending side of government, not the taxation side.
Redistribution depends on both taxation and spending, specifically whether the difference between the amount you pay in taxes and the amount you benefit from the spending is correlated with wealth and/or income. I had already argued that the increased taxes under a progressive tax policy are not balanced by an increase in benefits, which leaves as the only other option a situation where the difference is redistributed to others who pay less in taxes and yet benefit more from the spending.
As for insurance, if you buy a car that cost three times as much as I do, you would expect to pay three times the insurance, assume the other factors being the same. And yet, with taxes, at least in the US, it does not work that way.
The "taxes as insurance" analogy isn't a very good one—unless you mean "insurance" as in "protection racket", then it's fairly apt—but also consider this: if you can afford to lose the car, you don't need (comprehensive) insurance on it at all. You'll still need liability insurance, of course, but that's for a fixed amount and doesn't depend on the value of the vehicle (and if you're sufficiently wealthy you may be able to self-insure for liability as well).
If taxes were analogous to insurance then the rates would be regressive, not progressive. The wealthier you are the less protection you need, proportionally speaking. You have more to lose, but you can afford to lose more without serious hardship. The wealthy do not require proportionally greater levels of government services than the poor, and can easily afford to provide those services for themselves. The purpose of a progressive tax code is redistribution, not fairness.
I find it doubtful this is technically impossible as Apple are making out. ... This capability certainly exists in their test labs otherwise every update would be at risk of bricking phone on general release.
The risk isn't "bricking" the phone, just erasing the storage and starting over. Not a big deal, especially in a test lab environment.
This particular phone may be just barely old enough that the security features are implemented in software. Newer models, starting with the iPhone 5S, have already fixed that flaw by moving features like the minimum delay between attempts and auto-erasing the key after 10 failures into the Secure Enclave. Also, newer versions do not permit updates to be installed while the device is locked. If the phone were just one model newer then it would be demonstrably impossible for Apple to produce a software update to circumvent the security. As it is, they might be able to comply—but they shouldn't.
What this case is about what a third party can reasonably be ordered to do (without compensation?) to facilitate a legal search. This goes well beyond a landlord being ordered to unlock a back door. Or even allowing a wire tap to be installed on a phone line. My guess would be that assisting the FBI would probably take a few days and potentially disrupt Apple's iOS QA cycle for that long if they have to utilize in house resources.
Maybe longer since they essentially have to fork the iOS code base for this one device and then somehow isolate and target this one device for a software update. Oh and really trying hard not to brick the phone in the process. Not trivial, but certainly a somewhat borderline case considering the relatively vast resources of Apple.
All true, but I would argue that the fact that they're being asked to do something which they find morally repugnant, and the potential damage to their reputation, are greater considerations than the marginal costs in time, labor, and materials, which they would likely be (partly) compensated for in any event. Signing a firmware build is not just a technical measure to carry out a warrant; it's a statement by Apple that this build of the firmware has their seal of approval and is authorized to run on the iPhone. They have not even been accused of breaking any law, much less doing anything actually immoral, and yet they are being ordered by the court to provide false testimony under duress—all for the sake of gaining access to a dead suspect's work phone which probably doesn't contain any information pertinent to the investigation.
Well, that probably beats ROT-13, though I wouldn't care to stake anything important on it.
However, it is most definitely not a one-time pad. Despite being provably unbreakable when implemented correctly, OTP is perhaps the simplest of all encryption algorithms. There is no hashing, key-stretching, chaining, etc., just a message and a true-random pad of the same length. To encrypt or decrypt you just XOR the message and the pad.
The difficulty, of course, is in the key exchange. You have to arrange for both parties to securely share the same random pad, and any given bit of the pad can only be used to communicate one bit of message. Moreover, every pair of endpoints requires a different pad—there is no OTP equivalent of public-key cryptography.
This is where quantum cryptography really shines, as it employs quantum effects to negotiate random pads in real time between two endpoints without risk of undetected eavesdropping. The encryption algorithm in quantum cryptography is really just OTP; the novel part is the solution to the key exchange problem.
Actually, there is a hack for that. I saw a demonstration of a device that uses the USB interface to try a small number of PINs, then resets the device before it records the failures, then tries another small number of PINs, and so on.
This is why you should record the attempt first, before checking whether the PIN is valid. Resetting the device at any point in the authentication process before the attempt has been recorded should not yield any additional information about the PIN.
In this case, we have a lawful order from a court. Legally and morally, Apple should comply.
Legally, perhaps. The order may well be unlawful; compelling a third party not under investigation to produce brand-new software to circumvent existing security measures and sign it with their seal of approval as an approved software update is well outside the bounds of what these order are normally used for.
Morally, even if the order turns out to be legal, Apple should refuse. Regardless of its legality, the order itself is immoral and should not be followed. Following the order would be harmful to the security of every iPhone 5c owner, the vast majority of whom are innocent bystanders and not accused of any wrongdoing.