High speed rail is already close to too expensive to be practical. It is still a big question mark if those Maglev trains will ever be economically sound.
So if I understand the summary correctly (I give myself a 50/50 chance on this), they're basically sampling random noise off of a CCD and claim that eventually it will produce the Mona Lisa? A version of the million monkeys at typewriters producing Shakespeare?
This is really the key. You don't have to know the ins and outs of snapchat to teach Big O notation or data structures. That said, most primary and secondary CS teachers are way behind the times on lots of technology, and can often teach material that is no longer relevant. If they start talking about optimizing compute cycles (a topic that often comes up after Big O notation), then they're almost certainly going to be wrong now. CPU performance is dominated by cache misses, and organizing your data accesses can be far more important than reducing the total number of operations taken.
My parents did the same thing a couple of years ago. They needed to type up a letter on a new machine and didn't want to pay $$$$ for MS Office, so I suggested OpenOffice. An hour later they're calling saying that it doesn't work and that their machine is acting weird and ads are popping up constantly even when the machine is idle. Turns out they had downloaded it from god knows where and got the version bundled with an insane amount of crapware, instead of just downloading it from OpenOffice.org.
Now when I suggest software I always include the link to the correct download, because that's much faster than trying to remove a dozen different toolbars that are all fighting to take total control of the machine.
Never download software from one of those "Free Software Download" sites. They always bundle in crapware. Instead, track down the original author's homepage and try to download it from there. That greatly reduces the amount of crap you have to deal with.
Also, if you are forced to download from one of those sites, don't assume that just because you uncheck all of the crapware in the installer that it won't just go ahead and install it anyway, because it will. Basically, ask yourself if you really really need that app or if you could maybe find something else that does the same thing but is still supported. It's also a good idea to run whatever your favorite anti-spyware app is if you do have to install something like that.
You have to do more than overcome the internal resistance of the battery, you also have to supply energy faster than the natural discharge rate of the battery.
There are a dozen different memory technologies that "in 10 years time" will revolutionize everything. I'll believe it when I see it. Until this, this gets filed away with Bubble RAM and whatnot in the "will be nice if it ever pans out" file.
As usual the devil is in the details. Harvesting enough energy from the environment to run a very low power IC in short bursts to do simple sensor measurements? Completely feasible. Charging your cellphone from ambient radio waves? Complete bullcrap.
This was the dumbest thing. Not including the basic DNS functionality in the router advertisement--on a protocol utterly dependent on DNS because the addresses are so ugly--was a colossal blunder. Even then, stateless autoconfig has no mechanism to notify the DNS of the address it chose, so good luck populating a DNS server. Sure you don't need a hostname if you're purely a client, but that's far too narrow a view for how people actually use networks. I like a lot of what IPv6 does and think that in 10-15 years v4 will be a historic relic used in legacy networks and by old equipment, but it has some ugly warts right now.
I understand that the designers thought DNS was "just a service", but it's not. It a mapping tool for addresses at the IP layer. It's right there and more importantly, it has direct interaction with the IP layer. It should have been at the table during the IPv6 development.
Apple is in the same boat, but the new "superflat" UIs that are the current fad are horrible. It's rarely intuitive how anything works or what is or is not an interface element. I get that UI developers think buttons, switches, and sliders are ugly, but if you remove all of them it's really hard to use the device. People still aren't trained to swipe randomly all over the screen to try to figure out what mystery gesture does what, they just think the features are removed. Heaven forbid the UI tell you what gestures are available too, that would be cheating. Phones should be like puzzle boxes apparently.
The horribly useless comment system on Tumblr is his design I believe. If there was something before it, I shudder at the thought of what would be worse than the current system. It makes Slashdot Beta look like Slashdot Classic. I've found more useful information in Youtube comments than on Tumblr. Do we really need a whole line for every single person who ever liked or relinked or appreciated some post?
Depends. I can do an oil change in my truck with 5 quarts of cheap Costco oil in a half hour or so. My wife's car on the other hand demands 6 quarts of unicorn oil and it's basically impossible to save money over having someone else do it.
Even back in the day there weren't too many people who would fix ICs. Mostly just the guys who built them originally (radio enthusiasts for example). Electrical faults are really hard to find because you can't see them. I also can't fault kids for not wanting to work on modern electronics. Everything is so damn tiny now that you need to be a master surgeon to replace stuff. I replaced a cracked screen on my wife's iPhone 4S and that was a nightmare of tiny (0.7mm-1.2mm) screws of all different sizes and fragile tiny parts that had to be disassembled and reassembled in exactly the right way. Plus there was some glue. I'm sure a couple of the corner screws went back in cross threaded, and although the repair was successful I doubt I could do it again on that phone without stripping at least one or two of the screws.
I don't think the engineers are oblivious to the problem, at least most of the time. You don't decide to be a car engineer because you know nothing about cars. Sometimes you're given a crap sandwich and have to make it edible. Marketing says they want the new V6 in the compact so they can market it as a sport version, but the engine bay was originally designed for a tiny economy engine. And you can't change the dimensions because the bodywork people aren't looking at this model this year, so you have to find some place to cram all of the parts and you end up with craziness like batteries stuffed into windshield wiper reservoir spaces or oil filters that you have to remove a wheel to replace or engines where almost every repair starts with "remove the engine". And then 3/4 of the way through the project the marketing guy comes by and says that they've been telling everyone it's turbocharged, so could you add a turbocharger and oil intercooler while you're at it?
Wouldn't the criminals just create new wallets for each transaction? Blocking a wallet seems futile, they could have the money put of it and into a tumbler well before you got the info disseminated to all of the pools.
This is a huge factor. Theaters have been pushing ticked prices up much faster than the rate of inflation for at least a decade now. It's even worse if the movie is 3D, since they'll add $2 to each ticket to rent a pair of glasses for 90 minutes. It's really no wonder that ticket sales are down. The prices are not sustainable, yet I doubt I'll see any theater chains backing off of them anytime soon. Instead they're just going to cry about how home video is ruining their business, just like DVD and VCR before it.
The second approach runs into trouble if your tasks aren't independent. Parallel processing works great until you have to start synchronizing state. If one process stalls and the other processes are dependent on it for some data, then the other processes are going to stall anyway. In the real world, most problems are hard to separate cleanly--data dependencies are very very common. So there is a hidden cost to parallelism--the cost of synchronization between the threads, and the cost grows very fast as you add more threads. This is basically Linus's point: outside of specialized domains it's just not possible to cleanly break up most problems into more than just a handful of threads, so having a 1,000 core beast of a processor doesn't help. You would just have 990+ cores waiting on some other core to finish its job, all of the time. Plus there's the fact that debugging multithreaded programs is inherently more difficult than single threaded ones and that all of this is moot if you are I/O bound anyway.
About 3/4 of the way down the "article" they explained the vulnerability:
To control administrative access, Linux keeps a list of all the registered users on a machine, in a group typically known as “wheel,” who can be granted full root access (usually through the Unix sudo command).
A knowing attacker could get full root access by modifying the wheel group, either directly or by manipulating an adjoining program such as the Polkit graphical interface for setting user permissions, Alert Logic said.
This is patently stupid. Yes, if you give a badguy administrative access, bad things can happen--even if you use a fancy GUI to give the bad guy administrative access. The only thing that is even slightly newsworthy here is that maybe a novice admin won't understand the purpose of the wheel group and could be tricked into giving permissions, but there are a lot of ways you can trick a dumb admin, there's no need to single this one out.
Sony should say screw you to North Korea and release the entire movie for free on the internet. Make sure everybody has a chance to see it. Of course they won't because they still have to monetize it somehow, but it would be something to say "we're not going to give in".
The hosts file doesn't scale. It's only useful for LAN traffic and has a last ditch hack anymore. Especially since if a site is being attacked by a media cartel, they're probably going to have to switch servers a few times.
This will of course fail if the site you are trying to visit is virtually hosted. Even if you know the IP, you have to tell the proxy/load balancer what site you are trying to load.
Slashdot Mirror
High speed rail is already close to too expensive to be practical. It is still a big question mark if those Maglev trains will ever be economically sound.
So if I understand the summary correctly (I give myself a 50/50 chance on this), they're basically sampling random noise off of a CCD and claim that eventually it will produce the Mona Lisa? A version of the million monkeys at typewriters producing Shakespeare?
This is really the key. You don't have to know the ins and outs of snapchat to teach Big O notation or data structures. That said, most primary and secondary CS teachers are way behind the times on lots of technology, and can often teach material that is no longer relevant. If they start talking about optimizing compute cycles (a topic that often comes up after Big O notation), then they're almost certainly going to be wrong now. CPU performance is dominated by cache misses, and organizing your data accesses can be far more important than reducing the total number of operations taken.
My parents did the same thing a couple of years ago. They needed to type up a letter on a new machine and didn't want to pay $$$$ for MS Office, so I suggested OpenOffice. An hour later they're calling saying that it doesn't work and that their machine is acting weird and ads are popping up constantly even when the machine is idle. Turns out they had downloaded it from god knows where and got the version bundled with an insane amount of crapware, instead of just downloading it from OpenOffice.org.
Now when I suggest software I always include the link to the correct download, because that's much faster than trying to remove a dozen different toolbars that are all fighting to take total control of the machine.
Never download software from one of those "Free Software Download" sites. They always bundle in crapware. Instead, track down the original author's homepage and try to download it from there. That greatly reduces the amount of crap you have to deal with.
Also, if you are forced to download from one of those sites, don't assume that just because you uncheck all of the crapware in the installer that it won't just go ahead and install it anyway, because it will. Basically, ask yourself if you really really need that app or if you could maybe find something else that does the same thing but is still supported. It's also a good idea to run whatever your favorite anti-spyware app is if you do have to install something like that.
You have to do more than overcome the internal resistance of the battery, you also have to supply energy faster than the natural discharge rate of the battery.
There are a dozen different memory technologies that "in 10 years time" will revolutionize everything. I'll believe it when I see it. Until this, this gets filed away with Bubble RAM and whatnot in the "will be nice if it ever pans out" file.
As usual the devil is in the details. Harvesting enough energy from the environment to run a very low power IC in short bursts to do simple sensor measurements? Completely feasible. Charging your cellphone from ambient radio waves? Complete bullcrap.
This was the dumbest thing. Not including the basic DNS functionality in the router advertisement--on a protocol utterly dependent on DNS because the addresses are so ugly--was a colossal blunder. Even then, stateless autoconfig has no mechanism to notify the DNS of the address it chose, so good luck populating a DNS server. Sure you don't need a hostname if you're purely a client, but that's far too narrow a view for how people actually use networks. I like a lot of what IPv6 does and think that in 10-15 years v4 will be a historic relic used in legacy networks and by old equipment, but it has some ugly warts right now.
I understand that the designers thought DNS was "just a service", but it's not. It a mapping tool for addresses at the IP layer. It's right there and more importantly, it has direct interaction with the IP layer. It should have been at the table during the IPv6 development.
Apple is in the same boat, but the new "superflat" UIs that are the current fad are horrible. It's rarely intuitive how anything works or what is or is not an interface element. I get that UI developers think buttons, switches, and sliders are ugly, but if you remove all of them it's really hard to use the device. People still aren't trained to swipe randomly all over the screen to try to figure out what mystery gesture does what, they just think the features are removed. Heaven forbid the UI tell you what gestures are available too, that would be cheating. Phones should be like puzzle boxes apparently.
Cryptowall specifically overwrites all shadow copies of files.
The horribly useless comment system on Tumblr is his design I believe. If there was something before it, I shudder at the thought of what would be worse than the current system. It makes Slashdot Beta look like Slashdot Classic. I've found more useful information in Youtube comments than on Tumblr. Do we really need a whole line for every single person who ever liked or relinked or appreciated some post?
Depends. I can do an oil change in my truck with 5 quarts of cheap Costco oil in a half hour or so. My wife's car on the other hand demands 6 quarts of unicorn oil and it's basically impossible to save money over having someone else do it.
Even back in the day there weren't too many people who would fix ICs. Mostly just the guys who built them originally (radio enthusiasts for example). Electrical faults are really hard to find because you can't see them. I also can't fault kids for not wanting to work on modern electronics. Everything is so damn tiny now that you need to be a master surgeon to replace stuff. I replaced a cracked screen on my wife's iPhone 4S and that was a nightmare of tiny (0.7mm-1.2mm) screws of all different sizes and fragile tiny parts that had to be disassembled and reassembled in exactly the right way. Plus there was some glue. I'm sure a couple of the corner screws went back in cross threaded, and although the repair was successful I doubt I could do it again on that phone without stripping at least one or two of the screws.
I don't think the engineers are oblivious to the problem, at least most of the time. You don't decide to be a car engineer because you know nothing about cars. Sometimes you're given a crap sandwich and have to make it edible. Marketing says they want the new V6 in the compact so they can market it as a sport version, but the engine bay was originally designed for a tiny economy engine. And you can't change the dimensions because the bodywork people aren't looking at this model this year, so you have to find some place to cram all of the parts and you end up with craziness like batteries stuffed into windshield wiper reservoir spaces or oil filters that you have to remove a wheel to replace or engines where almost every repair starts with "remove the engine". And then 3/4 of the way through the project the marketing guy comes by and says that they've been telling everyone it's turbocharged, so could you add a turbocharger and oil intercooler while you're at it?
Wouldn't the criminals just create new wallets for each transaction? Blocking a wallet seems futile, they could have the money put of it and into a tumbler well before you got the info disseminated to all of the pools.
This is a huge factor. Theaters have been pushing ticked prices up much faster than the rate of inflation for at least a decade now. It's even worse if the movie is 3D, since they'll add $2 to each ticket to rent a pair of glasses for 90 minutes. It's really no wonder that ticket sales are down. The prices are not sustainable, yet I doubt I'll see any theater chains backing off of them anytime soon. Instead they're just going to cry about how home video is ruining their business, just like DVD and VCR before it.
The fact that we still don't have a long term solution for the waste is a concerning one. Yucca mountain was our best hope, and its dead.
The second approach runs into trouble if your tasks aren't independent. Parallel processing works great until you have to start synchronizing state. If one process stalls and the other processes are dependent on it for some data, then the other processes are going to stall anyway. In the real world, most problems are hard to separate cleanly--data dependencies are very very common. So there is a hidden cost to parallelism--the cost of synchronization between the threads, and the cost grows very fast as you add more threads. This is basically Linus's point: outside of specialized domains it's just not possible to cleanly break up most problems into more than just a handful of threads, so having a 1,000 core beast of a processor doesn't help. You would just have 990+ cores waiting on some other core to finish its job, all of the time. Plus there's the fact that debugging multithreaded programs is inherently more difficult than single threaded ones and that all of this is moot if you are I/O bound anyway.
Basically, the Seagate drive was $200 more expensive and about 20% faster than the WD drives. The WD Red and Green drives were basically identical.
This is patently stupid. Yes, if you give a badguy administrative access, bad things can happen--even if you use a fancy GUI to give the bad guy administrative access. The only thing that is even slightly newsworthy here is that maybe a novice admin won't understand the purpose of the wheel group and could be tricked into giving permissions, but there are a lot of ways you can trick a dumb admin, there's no need to single this one out.
What's so crazy about the idea that they lived out the rest of their lives with assumed identities?
Sony should say screw you to North Korea and release the entire movie for free on the internet. Make sure everybody has a chance to see it. Of course they won't because they still have to monetize it somehow, but it would be something to say "we're not going to give in".
The hosts file doesn't scale. It's only useful for LAN traffic and has a last ditch hack anymore. Especially since if a site is being attacked by a media cartel, they're probably going to have to switch servers a few times.
This will of course fail if the site you are trying to visit is virtually hosted. Even if you know the IP, you have to tell the proxy/load balancer what site you are trying to load.