Before Snowden we would have said the same thing about mass government surveillance.
Except that in this situation, we are not talking about surveillance, we are talking about bricking phones, which is quite the opposite. Surveillance is about gathering as much data as possible. Bricked phones don't gather much data. Plus, the kind of surveillance uncovered by Snowden is the of "spying" kind, where the goal is to make the targets unaware that they are being watched. In contrast, bricking phones is very obvious.
In Linux, there is no ABI. Drivers have to be accepted and included in the kernel source tree. Yes really. It's that well thought out.
This means that you have to have code review from the Linux kernel team. And you have to divulge any amateur or buggy code embodied in the source. Which may compromise the imaginary advantage your marketdroids think they have on other platforms.
That's linux stance, which is consistent with both the monolithic design and open source philosophy. However, this is not without problems. First problem is that some drivers can't reasonably be opensourced as they offer a real competitive advantage to the companies writing them. Or they may use licensed code that prevent them to go opensource. Second problem is that it relies on "kernel developers" or "the community". We shouldn't forget that these are made up of people with limited time and resources and chances are that they won't bother with your obscure long tail device* which mean that you'll have to constantly maintain your driver for every changing kernel version. Third problem is that stable linux distros are commonly build around a specific kernel version. If, for some reason, you need use hardware that is too recent you can't easily add support without changing the whole kernel.
Put in a 2 second delay between one road getting the red and the other getting the green. Done. Do people still rush lights? Yes they do....but.... a lot less collisions.
Well, where I live, it is already the case. I thought it was the same everywhere...
There is nothing special about positive and negative in formal proofs, in fact, using contraposition, they can be seen as equivalent (A->B !B->!A !A|B).
It only matters in law due to the presumption of innocence, and even then it isn't absolute, otherwise the judicial system would be pretty much powerless. The idea is that the accusation only has to prove that you are very likely to have committed the crime. After that it's up to you to prove that you didn't do it. Defense is a negative proof.
- Red light cameras are not at all red lights, only on random "profitable" red lights - There may or may not be signs on red light cameras, same with speed cameras - Rushing a red light is bad, but it is not always more dangerous than exceeding the speed limit. Crossing when the light just turned red or during traffic congestion is more of an annoyance than a real danger. And while you are clearly at fault if you get T-boned because you rushed a red light, it may not have been deadly it the other guy obeyed the speed limit. Excessive speed not only causes accidents, it also makes accidents worse even if you are not the one who caused it.
If you are a man and your have a women in your team, put her in front, whoever is the actual leader. Or maybe at least let your wife/GF/sister/whatever do the advertizing. I believe it should better if she is hot (but not slutty) as it should work on both men and women.
This is assuming of course that the study is unbiased and that gender really matters (i.e. : causation, not simple correlation).
Fax was much more popular in Japan than it was in the west so it's no wonder that it is still in use today. One of the reason for the popularity of fax, in addition to writing, are maps. Japanese addresses have no street names, they use a combination of district/block/house numbers which is very effective at losing people. As a result, it is common practice to send a map to mark a meeting place.
Wireless charging is not that good. It is energy inefficiant and you can't move your phone while you are charging. As for data transfer, wires are more secure and more reliable. Mag-safe type connectors are excellent for frying pots (that's were they were first used) and good for laptops but mobile devices are too light for mag-safe to offer a meaningful advantage.
I think they can get away with it using better materials (thicker gold plating maybe) and electrical protection built in the cable electronics (lightning cables are active). Remeber that apple doesn't have the same cost constraints as other manufacturers due to their ludicrous sale prices.
What you have to consider is that Apple devices are almost luxury items, which mean that they can get away using components that are too expensive for low-end devices. USB is designed to be cheap and good enough, otherwise it couldn't be "universal".
I thought that the 15-20 minute delay is simply because the cinavia signal is really low bandwidth (0.2bps according to wikipedia). As a result it takes a long time to check the key.
This study is interesting as it doesn't show that affirmative action itself has a positive effect. The simple knowledge that affirmative action is in place is sufficient, like some kind of placebo effect.
The idea is : women can win if they try but they don't try unless we tell them they have an unfair and in many cases unnecessary advantage.
Buffer in the main memory, software that optimize most-used code. It looks like an OS job for me, something that could be implemented in the linux kernel and benefit all CPUs, provided that you have the appropriate driver.
According to the paper, it looks like biggest novelty is... DRM. The optimizer code will be encrypted and will run in its own memory block, hidden from the OS. It will also make use of some special profiling instructions which could as well be accessible to the OS. Maybe they will but they say nothing about it.
Yes, auto-backup-restore from a central server is the obvious solution. However you have to do it properly, or else, it will become the weak point. You have to be careful of packet sniffing and man-in-the-middle attacks. Your server can be attacked too. And the more convinient you make your backups, the less secure they tend to be.
I think that the best compromise to turn on full disk encryption and that in case of anomaly (such as too many failed unlocks) the phone shuts down. Properly encrypted data are almost as good as a full (secure) wipe and better than an unsecure wipe.
Personally I didn't find a correlation between how I struggle writing code and the number of bugs I produce. I'm not sure about why but it's probably because I'm more careful with hard problems.
Anyways, I think the worst bug-generator is code that is hard to test, or not tested enough because of time constraints or poor management. This can explain the correlation : complex code (lots of inputs, lots of steps) tends to make programmers struggle and is hard to test. Also, programmers having a hard time tends to drive the project over budget, and when this happen, testing is commonly sacrified in a misguided attempt to recoup the goals.
I think that if we see developers struggling, instead of trying to "improve" the developers (the hard problem they are solving right now is actually what makes them better), it's better prepare to ramp up the testing budget, and maybe do a bit or code refactoring.
You have to balance things somehow. I'm not sure many people will want their phone to be wiped just because someone looks at it funny. If you make it easy to inadvertantly wipe data, you also need to have easy to access backups and these can be a security issue in their own right.
What's the difference between the NSA having 10 ways to hack into your computer vs having 100 ways ? The NSA can do whatever it wants in both cases. Except in the second case, there'll be less exloits available to the much more dangerous blackhats.
Why are blackhats more dangerous ? Because the NSA will "just" invade your privacy. Blackhats will steal your identity, ransom you hard drive, use your computer as a spambot and turn over your private data to anyone with money (this includes the NSA).
In fact it's not really obscure. One of the most used bus for critical data is ARINC429 and it's a well defined and very simple standard. If you somehow manage to connect a device directly to the bus wires (you can't do this with WiFi), it is relatively easy to inject whatever data you want. However, connected equipment will double check everything they receive, so you have to craft consistent data, a much harder task. And even then some devices have additional analog sensors so you'll need to fool these too. If data end up inconsistent (the most likely result of your hacking attempt), backups will kick in, because of course, these systems are redundant. And in the last resort, don't forget that there are still pilots on board. Also you can forget about typical vulnerabilities such as injection or buffer overflows. Avionic systems are too simple for this : data are not interpreted and all sizes are fixed and known in advance.
Sure, these systems are not 100% safe (they are "just" 99.9999999% safe). Sometimes a catastrophic bug may happen like with the Ariane 5 rocket. However, if you have some malicious intentions, attempting to hack the avionics it probably the least effective and most difficult way to cause harm.
Openness was never the business model of Nintendo. In fact not being open was a key to Nintendo success since the Famicom. Nintendo is all about good quality exclusive games.
By going open they will compete with the much more versatile Android and iOS devices and the much more powerful PCs and to a lesser extant Xbone and PS4. They simply can't win in the open market.
Free is not as good as before. Mostly because its competitors have adapted. They now offer similar prices and services via their low-cost labels. - Orange/SoSH is slightly more expensive but it has the best network and service - Bouygues/B&You use aggressive pricing and discounts. Beside 4G coverage, their service and network are rather poor. They are the most direct competitors of Free. - SFR/RED has a bit of a trouble keeping up with the competition as they don't offer same service as Orange nor the prices of Bouygues. They still have the second best network and may cover a few areas better than Orange. - Free has few offers but they are very good and cheap. However their network is the worst. They have a deal with Orange which allows them to have good coverage, however mobile data is severely throttled compared to Orange clients.
Xenophobes have the right to speak too, as long as they do it by the rules. If you deny the rights of a sizable part of the population to express themselves just because the majority disagrees with *some* of their ideas, it's not democracy.
Here is an simplified example of the problem : - 30% of the population are represented by a "far-right xenophobic party", will always vote against laws favoring immigrants - 60% are represented by a centrist party that have more nuanced ideas - 10% are represented by a far-left party that always favor immigrants In all cases, unreasonable demands made by the far-right and far-left party will never pass. Whether or not you give full power to the centrist party, the extremists will never get the required majority. Now imagine a controversial proposition favoring immigrants (maybe some kind of reverse discrimination). Let's say that 55% of the general population is against it and only 45% wants it. Normally, it should be rejected. However, if you give full power to the centrist party, it will be adopted. It is because the repartition will be as follows : - 30% are against in and are far-right - 25% are against it and are centrist - 35 % are for it and are centrist - 10 % are for it and are far-left
Giving governing responsibilities to small parties may even protect against extremes in some cases. Consider the following : same situation as before, except that the far-right party, though some clever political moves, manages to grab 25% of the centrist party, reaching 55% and winning the elections. The ideas of the people are the same, it's just that the far-right party is a bit less extreme than before. In a proportional system the remaining 45% are enough to keep it in check. Without it, the 30% original extremists would win as they are more numerous than the 25% centrists turned right-wing.
No. Last time I looked (by diffing the source code) AdBlock Edge was the same as AdBlock Plus without the "acceptable ads" checkbox. And when I look at the commits, almost all of them are backports from AdBlock Plus, no original development. Basically, what you get with AdBlock Edge is an outdated version of AdBlock Plus with less choice. Nothing wrong with forking ABP but at least do it for something useful.
And it's not like ABP tries to trick you. The "acceptable ads" option is clearly visible when you install the addon. If you know about ABE, you know about this option, so what the point ?
Imagine I am a provider of electricity, 20% of my energy comes from clean sources, 80% from dirty sources. I supply factories, datacenters, homes, etc... they all get the same electricity, after all, there is no such thing as clean electrons and dirty electrons. Now let say one of my client (that buys 6% of my production) says "I want 100% clean energy", what will I do ? Simple : I don't change anything in the way I produce electricity, I just say "OK, you now have 100% clean energy" and my other client now get 85% dirty energy.
If it is too obvious, there is always the option of splitting my company into two : one that only makes clean electricity, the other does does almost only dirty electricity. "Green" clients buy from my clean branch and others buy from my dirty branch, and if there is an excess of clean electricity, it is bought by my dirty branch.
The article show a clear correlation between the "iPhone slow" searches with the release dates. Only, without extra data, it doesn't mean anything. For example, what about the "iPhone" searches ? It's natural that when the new iPhone comes out, people will talk about iPhones, so all related queries, including "iphone slow" should raise. We need a baseline. Also "slow" doesn't always mean "slow performance". Notice the large peak corresponding the the 5S release, a phone that supports slow-motion video.
The article than compare it to the "Samsung Galaxy slow" queries (still without baseline). However, unlike the iPhone, the "Samsung Galaxy" brand encompass a large number of devices, from low-end to high-end, with releases all around the year. Yet again, such data are meaningless without further analysis.
Slashdot Mirror
So your situation is something you saw on 24?
Before Snowden we would have said the same thing about mass government surveillance.
Except that in this situation, we are not talking about surveillance, we are talking about bricking phones, which is quite the opposite.
Surveillance is about gathering as much data as possible. Bricked phones don't gather much data.
Plus, the kind of surveillance uncovered by Snowden is the of "spying" kind, where the goal is to make the targets unaware that they are being watched. In contrast, bricking phones is very obvious.
In Linux, there is no ABI. Drivers have to be accepted and included in the kernel source tree. Yes really. It's that well thought out.
This means that you have to have code review from the Linux kernel team. And you have to divulge any amateur or buggy code embodied in the source. Which may compromise the imaginary advantage your marketdroids think they have on other platforms.
That's linux stance, which is consistent with both the monolithic design and open source philosophy.
However, this is not without problems.
First problem is that some drivers can't reasonably be opensourced as they offer a real competitive advantage to the companies writing them. Or they may use licensed code that prevent them to go opensource.
Second problem is that it relies on "kernel developers" or "the community". We shouldn't forget that these are made up of people with limited time and resources and chances are that they won't bother with your obscure long tail device* which mean that you'll have to constantly maintain your driver for every changing kernel version.
Third problem is that stable linux distros are commonly build around a specific kernel version. If, for some reason, you need use hardware that is too recent you can't easily add support without changing the whole kernel.
(*) : no, I'm not talking about a black cat
Put in a 2 second delay between one road getting the red and the other getting the green. Done. Do people still rush lights? Yes they do....but.... a lot less collisions.
Well, where I live, it is already the case. I thought it was the same everywhere...
There is nothing special about positive and negative in formal proofs, in fact, using contraposition, they can be seen as equivalent (A->B !B->!A !A|B).
It only matters in law due to the presumption of innocence, and even then it isn't absolute, otherwise the judicial system would be pretty much powerless. The idea is that the accusation only has to prove that you are very likely to have committed the crime. After that it's up to you to prove that you didn't do it. Defense is a negative proof.
- Red light cameras are not at all red lights, only on random "profitable" red lights
- There may or may not be signs on red light cameras, same with speed cameras
- Rushing a red light is bad, but it is not always more dangerous than exceeding the speed limit. Crossing when the light just turned red or during traffic congestion is more of an annoyance than a real danger. And while you are clearly at fault if you get T-boned because you rushed a red light, it may not have been deadly it the other guy obeyed the speed limit. Excessive speed not only causes accidents, it also makes accidents worse even if you are not the one who caused it.
If you are a man and your have a women in your team, put her in front, whoever is the actual leader. Or maybe at least let your wife/GF/sister/whatever do the advertizing.
I believe it should better if she is hot (but not slutty) as it should work on both men and women.
This is assuming of course that the study is unbiased and that gender really matters (i.e. : causation, not simple correlation).
Fax was much more popular in Japan than it was in the west so it's no wonder that it is still in use today.
One of the reason for the popularity of fax, in addition to writing, are maps. Japanese addresses have no street names, they use a combination of district/block/house numbers which is very effective at losing people. As a result, it is common practice to send a map to mark a meeting place.
Wireless charging is not that good. It is energy inefficiant and you can't move your phone while you are charging. As for data transfer, wires are more secure and more reliable.
Mag-safe type connectors are excellent for frying pots (that's were they were first used) and good for laptops but mobile devices are too light for mag-safe to offer a meaningful advantage.
I think they can get away with it using better materials (thicker gold plating maybe) and electrical protection built in the cable electronics (lightning cables are active).
Remeber that apple doesn't have the same cost constraints as other manufacturers due to their ludicrous sale prices.
What you have to consider is that Apple devices are almost luxury items, which mean that they can get away using components that are too expensive for low-end devices.
USB is designed to be cheap and good enough, otherwise it couldn't be "universal".
I thought that the 15-20 minute delay is simply because the cinavia signal is really low bandwidth (0.2bps according to wikipedia). As a result it takes a long time to check the key.
This study is interesting as it doesn't show that affirmative action itself has a positive effect. The simple knowledge that affirmative action is in place is sufficient, like some kind of placebo effect.
The idea is : women can win if they try but they don't try unless we tell them they have an unfair and in many cases unnecessary advantage.
Buffer in the main memory, software that optimize most-used code. It looks like an OS job for me, something that could be implemented in the linux kernel and benefit all CPUs, provided that you have the appropriate driver.
According to the paper, it looks like biggest novelty is... DRM. The optimizer code will be encrypted and will run in its own memory block, hidden from the OS. It will also make use of some special profiling instructions which could as well be accessible to the OS. Maybe they will but they say nothing about it.
Yes, auto-backup-restore from a central server is the obvious solution.
However you have to do it properly, or else, it will become the weak point. You have to be careful of packet sniffing and man-in-the-middle attacks. Your server can be attacked too. And the more convinient you make your backups, the less secure they tend to be.
I think that the best compromise to turn on full disk encryption and that in case of anomaly (such as too many failed unlocks) the phone shuts down. Properly encrypted data are almost as good as a full (secure) wipe and better than an unsecure wipe.
Personally I didn't find a correlation between how I struggle writing code and the number of bugs I produce. I'm not sure about why but it's probably because I'm more careful with hard problems.
Anyways, I think the worst bug-generator is code that is hard to test, or not tested enough because of time constraints or poor management. This can explain the correlation : complex code (lots of inputs, lots of steps) tends to make programmers struggle and is hard to test. Also, programmers having a hard time tends to drive the project over budget, and when this happen, testing is commonly sacrified in a misguided attempt to recoup the goals.
I think that if we see developers struggling, instead of trying to "improve" the developers (the hard problem they are solving right now is actually what makes them better), it's better prepare to ramp up the testing budget, and maybe do a bit or code refactoring.
You have to balance things somehow. I'm not sure many people will want their phone to be wiped just because someone looks at it funny.
If you make it easy to inadvertantly wipe data, you also need to have easy to access backups and these can be a security issue in their own right.
What's the difference between the NSA having 10 ways to hack into your computer vs having 100 ways ?
The NSA can do whatever it wants in both cases. Except in the second case, there'll be less exloits available to the much more dangerous blackhats.
Why are blackhats more dangerous ? Because the NSA will "just" invade your privacy. Blackhats will steal your identity, ransom you hard drive, use your computer as a spambot and turn over your private data to anyone with money (this includes the NSA).
In fact it's not really obscure. One of the most used bus for critical data is ARINC429 and it's a well defined and very simple standard.
If you somehow manage to connect a device directly to the bus wires (you can't do this with WiFi), it is relatively easy to inject whatever data you want. However, connected equipment will double check everything they receive, so you have to craft consistent data, a much harder task. And even then some devices have additional analog sensors so you'll need to fool these too. If data end up inconsistent (the most likely result of your hacking attempt), backups will kick in, because of course, these systems are redundant. And in the last resort, don't forget that there are still pilots on board.
Also you can forget about typical vulnerabilities such as injection or buffer overflows. Avionic systems are too simple for this : data are not interpreted and all sizes are fixed and known in advance.
Sure, these systems are not 100% safe (they are "just" 99.9999999% safe). Sometimes a catastrophic bug may happen like with the Ariane 5 rocket. However, if you have some malicious intentions, attempting to hack the avionics it probably the least effective and most difficult way to cause harm.
Openness was never the business model of Nintendo. In fact not being open was a key to Nintendo success since the Famicom.
Nintendo is all about good quality exclusive games.
By going open they will compete with the much more versatile Android and iOS devices and the much more powerful PCs and to a lesser extant Xbone and PS4. They simply can't win in the open market.
Free is not as good as before. Mostly because its competitors have adapted. They now offer similar prices and services via their low-cost labels.
- Orange/SoSH is slightly more expensive but it has the best network and service
- Bouygues/B&You use aggressive pricing and discounts. Beside 4G coverage, their service and network are rather poor. They are the most direct competitors of Free.
- SFR/RED has a bit of a trouble keeping up with the competition as they don't offer same service as Orange nor the prices of Bouygues. They still have the second best network and may cover a few areas better than Orange.
- Free has few offers but they are very good and cheap. However their network is the worst. They have a deal with Orange which allows them to have good coverage, however mobile data is severely throttled compared to Orange clients.
Xenophobes have the right to speak too, as long as they do it by the rules.
If you deny the rights of a sizable part of the population to express themselves just because the majority disagrees with *some* of their ideas, it's not democracy.
Here is an simplified example of the problem :
- 30% of the population are represented by a "far-right xenophobic party", will always vote against laws favoring immigrants
- 60% are represented by a centrist party that have more nuanced ideas
- 10% are represented by a far-left party that always favor immigrants
In all cases, unreasonable demands made by the far-right and far-left party will never pass. Whether or not you give full power to the centrist party, the extremists will never get the required majority.
Now imagine a controversial proposition favoring immigrants (maybe some kind of reverse discrimination). Let's say that 55% of the general population is against it and only 45% wants it. Normally, it should be rejected. However, if you give full power to the centrist party, it will be adopted. It is because the repartition will be as follows :
- 30% are against in and are far-right
- 25% are against it and are centrist
- 35 % are for it and are centrist
- 10 % are for it and are far-left
Giving governing responsibilities to small parties may even protect against extremes in some cases. Consider the following : same situation as before, except that the far-right party, though some clever political moves, manages to grab 25% of the centrist party, reaching 55% and winning the elections. The ideas of the people are the same, it's just that the far-right party is a bit less extreme than before. In a proportional system the remaining 45% are enough to keep it in check. Without it, the 30% original extremists would win as they are more numerous than the 25% centrists turned right-wing.
No.
Last time I looked (by diffing the source code) AdBlock Edge was the same as AdBlock Plus without the "acceptable ads" checkbox. And when I look at the commits, almost all of them are backports from AdBlock Plus, no original development.
Basically, what you get with AdBlock Edge is an outdated version of AdBlock Plus with less choice. Nothing wrong with forking ABP but at least do it for something useful.
And it's not like ABP tries to trick you. The "acceptable ads" option is clearly visible when you install the addon. If you know about ABE, you know about this option, so what the point ?
It's free. If you don't like SecuROM, don't install the game.
It would be a valid argument if there were a clear mention of SecuROM before you install the game.
Imagine I am a provider of electricity, 20% of my energy comes from clean sources, 80% from dirty sources.
I supply factories, datacenters, homes, etc... they all get the same electricity, after all, there is no such thing as clean electrons and dirty electrons. Now let say one of my client (that buys 6% of my production) says "I want 100% clean energy", what will I do ?
Simple : I don't change anything in the way I produce electricity, I just say "OK, you now have 100% clean energy" and my other client now get 85% dirty energy.
If it is too obvious, there is always the option of splitting my company into two : one that only makes clean electricity, the other does does almost only dirty electricity. "Green" clients buy from my clean branch and others buy from my dirty branch, and if there is an excess of clean electricity, it is bought by my dirty branch.
The article show a clear correlation between the "iPhone slow" searches with the release dates. Only, without extra data, it doesn't mean anything. For example, what about the "iPhone" searches ? It's natural that when the new iPhone comes out, people will talk about iPhones, so all related queries, including "iphone slow" should raise. We need a baseline.
Also "slow" doesn't always mean "slow performance". Notice the large peak corresponding the the 5S release, a phone that supports slow-motion video.
The article than compare it to the "Samsung Galaxy slow" queries (still without baseline). However, unlike the iPhone, the "Samsung Galaxy" brand encompass a large number of devices, from low-end to high-end, with releases all around the year. Yet again, such data are meaningless without further analysis.