There's a fairly common bug in a lot of password systems (I first encountered it locking myself out of my Psion Series 3, but it was also in macOS for a long time) where you can enter arbitrary ASCII, 8-bit, or even unicode characters when you set the password, but not when you enter it to log in. Only try it with systems that have a good password reset mechanism!
Second, forget about it all you people with your **genius** schemes for generating unique 8-11 character passwords. Congratulations, you've just been hacked. Look up rainbow tables, people!
If you have upper- and lower-case letters, numbers, and symbols then each character is one from a set of 80, so a random 8-character password from this set contains 50 bits of entropy (2^50 possible combinations). To store all such passwords in a rainbow table would require 2^54 bytes (8 petabytes) of storage. I doubt that most hackers have that much space.
A case insensitive 8-character password, in contrast, has just under 38 bits of entropy, so it is quite feasible to compute a rainbow table. Mixing cases alone takes this up to 45 bits, which means that you'll need around half a petabyte for the rainbow table.
If you're using a salted hash to store the password, then the rainbow table needs to be computed for each salt (and if you're sensible, you'll use a different salt for each password, so you need a different rainbow table per password, not per password db). You're better off brute forcing it than storing the rainbow table. A modern GPU can manage about 20,000,000,000 hashes per second, so can search a 34-bit key space per second. 45 bit of entropy gives you a search space that takes about half an hour of GPU time. 50 bits gives you 18 hours. An 11-character password will give you 69 bits of entropy (and a rainbow table that most filesystems can't store, though ZFS can if you can afford enough disks), and will take about 1,000 years to brute force with a single GPU (though with 10,000 GPUs you can do it in a reasonable amount of time). A 10-character password gives you 63 bits, which takes about 17 GPU years to crack and is still probably beyond the capabilities of anyone other than a nation-state adversary.
Apple released the G5 iMac - a wall-mountable computer embedded in a display with a camera and network interface built in - on the 20th anniversary of their 1984 superbowl commercial, which ran with the tagline 'why 1984 won't be like 1984'. Apparently the reason was that it will take 20 years to get it into production.
Samsung TVs are quite popular. It's likely that they are in sensitive places, like meeting rooms of US corporations, hospitals, newsrooms etc.
It's worse than that. These TVs don't just end up there, they're actively marketed at these places because they can install various video conferencing apps and avoid the need to have a separate computer to control the video conferencing system.
And you can do that, as long as there's a set of wheels where you land (not necessarily at an airfield, as this uses a quadcopter for VTOL). The target for this is large autonomous networks of vehicles - you'll have your own pod and the system will route wheels / copters to you as needed. For example, using a copter to hop across a river, rather than driving a few miles to a bridge and then back again.
The concept works around some of that by having a passenger pod that is carried by either a quadcopter or a wheeled frame. Things like airbags and roll cages need to be in the pod, but most of the ground engine weight, crash bars, and fuel can remain on the ground.
Additionally, ejection systems are dangerous. The ejector seat mechanism compresses your spine to the extent that people who have ejected from a plane are measurably shorter than prior to ejection. If you're a healthy adult in good physical condition (i.e. the sort of person who is allowed to fly fighter jets) then you can do it a few times and survive (whereas you probably can't explode in a fireball a few times and survive, so ejecting is a better option), but if you eject more than a few times you'll be grounded on medical reasons. If you do the same with a typical commercial aircraft passenger, there's a reasonable chance that they'll die, whereas many commercial aircraft crashes are survivable.
As to just having a parachute, you'd probably only be able to jump from the rear exits without being sucked into the engines. You can't jump without oxygen until the plane is a lot lower than its cruising altitude, and if it can get down low enough to jump and stay there long enough to get 300 people out then it's probably able to manage a survivable landing. Landing from a parachute jump isn't that difficult, but generally requires a little bit of practice - at least some passengers wouldn't be able to do it. The bit after landing is also difficult, as you have to disconnect the chute quickly to avoid being dragged along and you must remember to disconnect the chest straps before the leg straps or you'll end up being strangled. And given that someone always panics and inflates their life jacket inside the plane, in spite of being repeatedly told not to, in simulated crashes, what's the bet that someone wouldn't pull the chord on their parachute by accident (actually surprisingly easy to do when putting the chute on) on the plane and kill / injure other passengers (those springs contains a lot of energy).
property taxes on the increased value of the house?
Property taxes here are assessed in bands and aren't revalued very often, so it's really only the difference between the initial purchase prices. That difference between bands for this kind of difference is pretty small - far less than the difference between mortgage payments.
Increased home owner's insurance for increased value of the house?
Contents insurance was lower than in the place I was renting, because the doors and windows were more secure and it was in a lower crime area. Building insurance costs don't vary very much with property value. I'm now in a house valued at about 4-5 times the current value first house I bought and property insurance prices are about 50% higher.
It's not just cheap labour, it's also often weaker pollution regulation. Dumping the waste from your factory in the local river can dramatically cut the cost of production compared with having to collect and process the same waste. That's been almost as big a driver for moving production to China, India, Africa, and so on as the cheaper labour. It's now harder in China, as they're starting to tighten up pollution laws and have executed a couple of officials for taking bribes to overlook polluting factories.
Bike lanes in Austin would need to be covered and air conditioned before I could imagine anyone using them. Last time I was there was spring and it was too hot to comfortably cycle by about 6:30am (a time I normally regard as fictional, but yay, jetlag). I'm told that in the summer it's a lot hotter. Maybe there are a couple of weeks of good cycling weather each year...
I did similar calculations when I was buying my first house: there were some houses a couple of miles out of town that were quite a bit cheaper, but would end up costing a lot more over a few years. If anything, he's being optimistic, because he's not factoring in the appreciation in the house. He does make the point that you can afford a much more expensive house if you're not commuting, but doesn't add in the fact that even a 1% appreciation per year in the value of that house over the time that you're living in it will come to around a third of the amount that he says that you save by commuting: 3% appreciation will more than double the savings. He also doesn't factor in things like the fact that living out of town also increases the cost of getting to shops or the fact that you need to get a taxi (or persuade your partner to be a designated driver) if you want to go out for drinks.
Which is made worse by the crazy zoning system that American cities seem to use, designed to ensure that places where people work, places where people shop or go for recreation, and places where people live are all as far away from each other as possible. When I first played SimCity, I thought it was absurd: no one would be stupid enough to design a city like that. Then I visited the US for the first time.
CalTrain is too infrequent. Even at peak times it is only about every half hour and those trains often don't stop at every stop. It's okay as a commuter train, but it's a bit painful for anything else. Good mass transit systems have a train every 2-5 minutes, so you hardly spend any time waiting. The BART manages every 10-20 minutes, depending on line and time of day. CalTrain can't decide whether it's a mass-transit system or a conventional rail line and manages to combine the worst aspects of both.
It's not a new thought experiment though. It's something that's been discussed, and even worked on, for decades. Some of us have been working in this space for a lot of that time. It's like saying 'imagine that all fabric had the same tensile strength'.
The downside is that Qt apps on macOS are hideous. It took them 10 years to get the key bindings for text fields to behave the same way as the rest of the system. They still end up with modal interfaces that don't fit at all with the rest of the system, use tabbed interfaces that don't fit with anything else on the system, and so on. They look and feel nothing like other macOS applications.
The grandparent brought up Java, not me. There are a lot of attempts at cross-platform UIs, including ones that only provide source compatibility and require that you recompile for each target. They all suck for the reasons that I highlighted. The only good cross-platform software has a shared core of code and then customises the UI for each target platform. In a lot of desktop software, the UI is a sufficiently large part of the total codebase that this amounts to an almost total rewrite.
In general, people who look at child pornography are people who have a sexual interest in children. And if you're trying to find people who are sexually abusing children then finding people with a sexual interest in children is a great way to start.
By the same argument, anyone who looks at porn involving adults is a potential rapist. It's pretty obvious that anyone who sexually abuses children is going to enjoy child pornography (though it's not clear that they're going to successfully find any). It's far less obvious that child pornography is some kind of gateway to child abuse, especially given that the vast majority of cases of child abuse are by the child's own parents.
But I doubt they'd be very interested in the downloaders if they didn't have a huge overlap with abusers.
Why? Both groups are about as unpopular in the media, one is a lot harder to catch. If I were setting priorities in a highly politicised law enforcement agency, I know which group I'd target.
And Java is also a good example of why it's a terrible idea. If your software runs on all platforms then it's limited to supporting the intersection of all of the features that those platforms support. For GUI software (if you want to be running exactly the same software on all platforms) it's also limited to respecting the human interface guidelines of at most one platform. This makes it particularly jarring to use on others. It's easy to fix the superficial things (dialog box buttons in the native order, menu in the correct location, and so on), but a lot of more subtle things (modal vs non-modal dialog boxes, apply settings as they're changed vs okay / cancel buttons and so on) are much harder to automatically translate.
Machines are not free and including the capital amortised over their lifetime they may cost several dollars per hour. They do keep getting cheaper though, so there was always going to be a point at which they'd become cheaper than humans for any given task, once they can be built to do that task at all. Raising the minimum wage may make that happen sooner, but probably not by more than a couple of years.
Slashdot Mirror
There's a fairly common bug in a lot of password systems (I first encountered it locking myself out of my Psion Series 3, but it was also in macOS for a long time) where you can enter arbitrary ASCII, 8-bit, or even unicode characters when you set the password, but not when you enter it to log in. Only try it with systems that have a good password reset mechanism!
Not me (and I'm not the grandparent), but some of my colleagues would say that 10MB is bloatware for a server system.
Second, forget about it all you people with your **genius** schemes for generating unique 8-11 character passwords. Congratulations, you've just been hacked. Look up rainbow tables, people!
If you have upper- and lower-case letters, numbers, and symbols then each character is one from a set of 80, so a random 8-character password from this set contains 50 bits of entropy (2^50 possible combinations). To store all such passwords in a rainbow table would require 2^54 bytes (8 petabytes) of storage. I doubt that most hackers have that much space.
A case insensitive 8-character password, in contrast, has just under 38 bits of entropy, so it is quite feasible to compute a rainbow table. Mixing cases alone takes this up to 45 bits, which means that you'll need around half a petabyte for the rainbow table.
If you're using a salted hash to store the password, then the rainbow table needs to be computed for each salt (and if you're sensible, you'll use a different salt for each password, so you need a different rainbow table per password, not per password db). You're better off brute forcing it than storing the rainbow table. A modern GPU can manage about 20,000,000,000 hashes per second, so can search a 34-bit key space per second. 45 bit of entropy gives you a search space that takes about half an hour of GPU time. 50 bits gives you 18 hours. An 11-character password will give you 69 bits of entropy (and a rainbow table that most filesystems can't store, though ZFS can if you can afford enough disks), and will take about 1,000 years to brute force with a single GPU (though with 10,000 GPUs you can do it in a reasonable amount of time). A 10-character password gives you 63 bits, which takes about 17 GPU years to crack and is still probably beyond the capabilities of anyone other than a nation-state adversary.
Apple released the G5 iMac - a wall-mountable computer embedded in a display with a camera and network interface built in - on the 20th anniversary of their 1984 superbowl commercial, which ran with the tagline 'why 1984 won't be like 1984'. Apparently the reason was that it will take 20 years to get it into production.
Samsung TVs are quite popular. It's likely that they are in sensitive places, like meeting rooms of US corporations, hospitals, newsrooms etc.
It's worse than that. These TVs don't just end up there, they're actively marketed at these places because they can install various video conferencing apps and avoid the need to have a separate computer to control the video conferencing system.
Bobs hacked Samsung TV is not a national security issue
It is when Bob is the son of a general and the television is used to eavesdrop on conversations of classified material in his house.
And you can do that, as long as there's a set of wheels where you land (not necessarily at an airfield, as this uses a quadcopter for VTOL). The target for this is large autonomous networks of vehicles - you'll have your own pod and the system will route wheels / copters to you as needed. For example, using a copter to hop across a river, rather than driving a few miles to a bridge and then back again.
The concept works around some of that by having a passenger pod that is carried by either a quadcopter or a wheeled frame. Things like airbags and roll cages need to be in the pod, but most of the ground engine weight, crash bars, and fuel can remain on the ground.
Additionally, ejection systems are dangerous. The ejector seat mechanism compresses your spine to the extent that people who have ejected from a plane are measurably shorter than prior to ejection. If you're a healthy adult in good physical condition (i.e. the sort of person who is allowed to fly fighter jets) then you can do it a few times and survive (whereas you probably can't explode in a fireball a few times and survive, so ejecting is a better option), but if you eject more than a few times you'll be grounded on medical reasons. If you do the same with a typical commercial aircraft passenger, there's a reasonable chance that they'll die, whereas many commercial aircraft crashes are survivable.
As to just having a parachute, you'd probably only be able to jump from the rear exits without being sucked into the engines. You can't jump without oxygen until the plane is a lot lower than its cruising altitude, and if it can get down low enough to jump and stay there long enough to get 300 people out then it's probably able to manage a survivable landing. Landing from a parachute jump isn't that difficult, but generally requires a little bit of practice - at least some passengers wouldn't be able to do it. The bit after landing is also difficult, as you have to disconnect the chute quickly to avoid being dragged along and you must remember to disconnect the chest straps before the leg straps or you'll end up being strangled. And given that someone always panics and inflates their life jacket inside the plane, in spite of being repeatedly told not to, in simulated crashes, what's the bet that someone wouldn't pull the chord on their parachute by accident (actually surprisingly easy to do when putting the chute on) on the plane and kill / injure other passengers (those springs contains a lot of energy).
No. No it isn't.
property taxes on the increased value of the house?
Property taxes here are assessed in bands and aren't revalued very often, so it's really only the difference between the initial purchase prices. That difference between bands for this kind of difference is pretty small - far less than the difference between mortgage payments.
Increased home owner's insurance for increased value of the house?
Contents insurance was lower than in the place I was renting, because the doors and windows were more secure and it was in a lower crime area. Building insurance costs don't vary very much with property value. I'm now in a house valued at about 4-5 times the current value first house I bought and property insurance prices are about 50% higher.
It's not just cheap labour, it's also often weaker pollution regulation. Dumping the waste from your factory in the local river can dramatically cut the cost of production compared with having to collect and process the same waste. That's been almost as big a driver for moving production to China, India, Africa, and so on as the cheaper labour. It's now harder in China, as they're starting to tighten up pollution laws and have executed a couple of officials for taking bribes to overlook polluting factories.
Bike lanes in Austin would need to be covered and air conditioned before I could imagine anyone using them. Last time I was there was spring and it was too hot to comfortably cycle by about 6:30am (a time I normally regard as fictional, but yay, jetlag). I'm told that in the summer it's a lot hotter. Maybe there are a couple of weeks of good cycling weather each year...
I did similar calculations when I was buying my first house: there were some houses a couple of miles out of town that were quite a bit cheaper, but would end up costing a lot more over a few years. If anything, he's being optimistic, because he's not factoring in the appreciation in the house. He does make the point that you can afford a much more expensive house if you're not commuting, but doesn't add in the fact that even a 1% appreciation per year in the value of that house over the time that you're living in it will come to around a third of the amount that he says that you save by commuting: 3% appreciation will more than double the savings. He also doesn't factor in things like the fact that living out of town also increases the cost of getting to shops or the fact that you need to get a taxi (or persuade your partner to be a designated driver) if you want to go out for drinks.
Which is made worse by the crazy zoning system that American cities seem to use, designed to ensure that places where people work, places where people shop or go for recreation, and places where people live are all as far away from each other as possible. When I first played SimCity, I thought it was absurd: no one would be stupid enough to design a city like that. Then I visited the US for the first time.
CalTrain is too infrequent. Even at peak times it is only about every half hour and those trains often don't stop at every stop. It's okay as a commuter train, but it's a bit painful for anything else. Good mass transit systems have a train every 2-5 minutes, so you hardly spend any time waiting. The BART manages every 10-20 minutes, depending on line and time of day. CalTrain can't decide whether it's a mass-transit system or a conventional rail line and manages to combine the worst aspects of both.
It's not a new thought experiment though. It's something that's been discussed, and even worked on, for decades. Some of us have been working in this space for a lot of that time. It's like saying 'imagine that all fabric had the same tensile strength'.
Block chains and Merkel trees have a lot in common.
Which is precisely why cross-platform GUI applications don't work as a concept.
The downside is that Qt apps on macOS are hideous. It took them 10 years to get the key bindings for text fields to behave the same way as the rest of the system. They still end up with modal interfaces that don't fit at all with the rest of the system, use tabbed interfaces that don't fit with anything else on the system, and so on. They look and feel nothing like other macOS applications.
The grandparent brought up Java, not me. There are a lot of attempts at cross-platform UIs, including ones that only provide source compatibility and require that you recompile for each target. They all suck for the reasons that I highlighted. The only good cross-platform software has a shared core of code and then customises the UI for each target platform. In a lot of desktop software, the UI is a sufficiently large part of the total codebase that this amounts to an almost total rewrite.
In general, people who look at child pornography are people who have a sexual interest in children. And if you're trying to find people who are sexually abusing children then finding people with a sexual interest in children is a great way to start.
By the same argument, anyone who looks at porn involving adults is a potential rapist. It's pretty obvious that anyone who sexually abuses children is going to enjoy child pornography (though it's not clear that they're going to successfully find any). It's far less obvious that child pornography is some kind of gateway to child abuse, especially given that the vast majority of cases of child abuse are by the child's own parents.
But I doubt they'd be very interested in the downloaders if they didn't have a huge overlap with abusers.
Why? Both groups are about as unpopular in the media, one is a lot harder to catch. If I were setting priorities in a highly politicised law enforcement agency, I know which group I'd target.
They care about locking people up so much that they're willing to drop a case rather than present evidence?
And Java is also a good example of why it's a terrible idea. If your software runs on all platforms then it's limited to supporting the intersection of all of the features that those platforms support. For GUI software (if you want to be running exactly the same software on all platforms) it's also limited to respecting the human interface guidelines of at most one platform. This makes it particularly jarring to use on others. It's easy to fix the superficial things (dialog box buttons in the native order, menu in the correct location, and so on), but a lot of more subtle things (modal vs non-modal dialog boxes, apply settings as they're changed vs okay / cancel buttons and so on) are much harder to automatically translate.
Machines are not free and including the capital amortised over their lifetime they may cost several dollars per hour. They do keep getting cheaper though, so there was always going to be a point at which they'd become cheaper than humans for any given task, once they can be built to do that task at all. Raising the minimum wage may make that happen sooner, but probably not by more than a couple of years.