Disney has an entire division devoted to cranking out crap sequels to hits. They're responsible for Cinderella 2 and 3, Bambi 2, Pocohontas 2, Mulan 2, Tarzan 2, The Lion King 1.5 and 2, The Jungle Book 2, Lilo and Stitch 2, and a host of others, most of which can be found wherever used DVDs are sold. So grinding out Star Wars 7 is in line with the established Disney production pipeline.
Yes, Star Wars 7 is nominally a live action film, but today that's just a few principal characters on top of CG animation. Most of the pixels come from the animation teams.
Right. Just like commercial air travel, elevators, and escalators. Which is the whole point.
This will be just fine with the trucking industry. The auto industry can deal with "boring" by putting in more cupholders, faux-leather upholstery, and infotainment systems.
The US areas that are in trouble are mostly the Gulf coast, especially the Mississippi River flood plain and Florida. Florida is just barely above sea level now, and is very flat.
Slight rises in sea level cause problems all along the Mississippi. Hurricane and storm driven flooding are already getting worse.
The West Coast isn't so bad off, because there are cliffs along most of it. SF, LA and San Diego do have low spots, but they're a few miles long, and seawalls could be built. It might be necessary to dam the SF bay, with something like the Thames Barrage at the Golden Gate.
There's a whole cult of "vertical farming". With LED lighting, it's much more cost-effective. Not only is the power consumption down, but the plants can be racked much more closely without overheating. Phillips has some special-purpose farming LEDs with spectra chosen for growing specific crops.
So far, most of the enthusiasm for this comes from the organic "farmer's market" crowd, not production-scale farmers.
Too many web sites which run ads are buying them through a chain of multiple resellers. Under current law, the web site running the ad can usually disclaim responsibility for hostile ads. That may change. The article is about testimony before the U.S. Senate's committee on homeland security.
The site that displays the ads should be held responsible. Sites which run ads would then need to protect themselves by legal and technical means. For example, if you run ads on your site, your contract with the advertising provider should provide that they will indemify and defend you should a bad ad get through.
The PS3's Cell processor is, in simplified terms, a general purpose CPU and six special purpose coprocessors.
The "coprocessors" are decent general-purpose computers. The problem is that they only have 128K of RAM each. They can access main memory, with high latency but good bandwidth, through a DMA mechanism. (There's also a relatively conventional NVidia GPU on the back end, so the GPU part of the job isn't the big difference. This isn't about shaders.)
That 128K of RAM is too small for a frame. Too small for a level. Too small for any big part of game state. PS3 programming thus consists in turning a problem into some form of streaming process, where data is pumped into Cell processors, processed a bit, and pumped back out. This is a signal processing architecture. It's great for audio. Sucks for everything else.
The PS3 was a "build it and they will come" architecture. It was cheap to make, but large numbers of smart people spent years trying to figure out how to use the thing properly. (Sony basically gutted their US R&D group to beat on that problem.) In practice, a lot of games did most of their work in the main CPU (a MIPS machine) and the GPU, using the Cell processors only for audio, fire and explosions, particle systems, and other tasks that didn't have a lot of interconnected state.
No, that's an ad for "Tile". Range 50 to 150 feet to nearest iPhone with the Tile app. Lifespan of one year.
It's another one of those "preorder" scams - pay now, get delivery someday, maybe. "With your help, 49,586 backers preordered Tiles totaling $2,681,297." They're vague about how many were actually delivered.
This is the kind of anti-competitive behavior that gets companies in trouble and causes regulatory crackdowns. Phone companies that make it hard to switch carriers. domain registrars that make it hard to switch registrars, and banks which make it hard to switch banks have all gotten in trouble for this.
If this is just being publicly announced now, it was probably known to NSA, the GRU, and the MSS years ago. The superpower security agencies put substantial resources into cryptanalytic number theory.
Early public key cryptosystems used the knapsack problem. That turned out to have reasonably easy solutions. Factoring products of two big primes may be all we have left. That's believed to be hard, but there's no proof of a lower bound on it.
If only. IBM mainframes have had an IOMMU since the IBM 360/67. On IBM mainframes, channels work through the memory protection system, and driver programs can't get a device to read or write outside the program's own space.
PCs and their descendants do not have that. In the early days, MMUs took too many transistors. The Intel line didn't have an CPU-side MMU until the 386 (when 32 bits came in) and Microsoft didn't use the MMU until Windows NT came out in 1993. So there wasn't much call for an IO-side MMU, until it was too late. There are some memory-mapping gimmicks on I/O buses today, but they're mostly there to deal with 32 vs. 64 bit addressing problems.
Peripherals in the PC world have been able to blither all over memory since the earliest days. It's gotten a little better; at least USB devices can't do that. That's important, because, in practice, attacks via USB are common. The buses that offer full memory access are usually internal to the machine, so plugging in a hostile device isn't easy. There are exceptions, including PCMCIA and ExpressCard, which do expose a bus. Exposing a bus on a cable is a very bad idea in today's hostile environment.
Manufacturing auto batteries should be a low-margin business. They're a commodity. Others can enter the business. Over time, margins will decrease. There's not much brand value. (Who made the battery in your phone?)
The article doesn't give the revision history of how that code got there. Who put it there, when did they put it there, and what did the code look like before and after they put it there. We need the names!
Most of the "patent troll" problem comes from three law firms. They're the only ones who've sent out more than two demand letters, according to the EFF.
Read "American Drive", by someone who did a startup in Detroit. His approach was to buy a failing GM axle plant cheap using money from a private equity firm, kick out the union, cut wages over 50%, put in some decent machinery, and make big bucks. That's a Detroit startup for you.
The most amusing part is how they dealt with the crack house across the street from their parking lot. They weren't getting much help from the Detroit cops. So they put stadium-sized lights on the light poles in their parking lot and aimed them all directly at the crack house. When those were switched on for the first time, it was like spraying an ant nest. People ran from the house. The crack house went out of business after a few weeks under the lights.
Parking garages, though... Lighted arrows and lines to direct people to empty slots might be useful. It could be useful for intersections which have highly variable traffic patterns, where adding additional turn lanes dynamically is useful. Stadiums often do that, with a small army of people moving traffic cones around.
Solar powered snow melting seems unlikely to work. If you really need snow melting, the power requirements are huge. The cutting edge of technology there is induction heating of snow in railroad switches. Many railroads in snowy areas heat their switches. But nobody heats the entire track.
Ah, startup battery companies. There have been so many.
I have no idea if this will work. But the company seems flaky. Their "path to market" says "This facility will allow Power Japan Plus to meet demand for specialty energy storage markets such as medical devices and satellites." In other words, they're nowhere near making auto batteries.
What these guys need to do is build some prototype batteries, put them in a car, and drive on battery power from Tokyo to Osaka without a recharge. That's a little beyond what a Tesla can do. If they can do that, people will listen.
Thunderbolt isn't going to replace USB in all cases, but Thunderbolt isn't about the speed. It's about the protocol. Thunderbolt is basically PCI-E over a wire.
Bad idea for security reasons. Any device plugged in can read and write memory. That's not a good thing. At least with USB, it's just packets to and from the driver.
FireWire had the same problem. Most FireWire PC interfaces allowed limiting the hardware capability to accept packets that read and wrote memory. (There were address limit registers. The default settings for Linux left memory wide open to FireWire attack. (Under Linux, all of memory was open on 32-bit systems, but because this was a bug, not a feature, only the first 4MB was open on 64-bit systems. I once reported this as a Linux kernel bug. There were people who didn't want it fixed because they were using it for kernel debugging.)
Being so close to water and all, the entire idea of an e-reader for naval sailors is preposterous.
This is for submariners. If they encounter water, they have much worse problems than e-reader failure. It's just for recreational reading, anyway; it's not for storing manuals. What the Navy wants is a device which absolutely cannot emit any kind of signal and possibly give away a sub's location.
It might even be waterproof. There are waterproof e-readers. The Navy tends to waterproof everything to at least the "resists high humidity and salt spray" level. That's so routine it's not usually mentioned. (There's no real problem waterproofing electronics. More mobile phones should be waterproof. Some already are; I have one.)
Tell a few of their customers to ship back items which appear to have been tampered with, and compare them at your end. That's appropriate tech support. You have no idea who's doing the tampering or why, and it's worth finding out.
The problems had nothing to do with the use of thorium fuel. It had everything to do with a badly designed cooling system that used He instead of water.
The Ft. St. Vrain story is rather sad. The plant had a large number of minor problems that made it too expensive to run. It was converted to natural gas.
Every reactor design which had something complicated happening within the radioactive parts of the system has been a commercial failure. Standard boiling-water reactors and pressurized-water reactors are very simple both mechanically and chemically inside the reactor vessel. All the complexity is outside, where it can be fixed if necessary.
Sodium-cooled reactors have sodium fires. Pebble-bed reactors have jams. (There's a prototype in Germany that's so jammed it can't be decommissioned.) Helium-cooled reactors have leaks. Reactors which require an adjacent chemical processing plant have all the problems of a chemical plant for radioactive materials. Anything which goes wrong in the radioactive part of the system is a huge deal to fix. The history of exotic reactor designs is not good. Many of the exotic ideas have been funded and built, but the results are not impressive.
Meanwhile, boring old BWR and PWR reactors have a long life and good uptime.
I don't use any Google account services. My mail goes to an IMAP server with spam filtering. The Linux desktop, the Windows 7 desktop, and the Windows laptop all run Thunderbird. The Android smartphone, which does not have a Google account, has an IMAP client. All devices sync mail through IMAP. Works fine.
Multidimensional arrays in C and C++ can be iterated over efficiently along any axis of your choosing. You simply choose which axis when you create the array.
That has to have been written by someone who doesn't do much matrix work. A basic matrix multiply iterates in row order for one matrix and column order for the other.
That's the problem. There's no one way to represent a multidimensional array in C++. There are many ways. Which means math libraries using different ones are incompatible with each other. The last time I did a big number-crunching job in C++, I had four different array representations forced on me by different libraries.
Because the compiler has no clue what those array libraries are doing, you don't get basic loop optimizations that FORTRAN has had for 50 years.
Slashdot Mirror
Disney has an entire division devoted to cranking out crap sequels to hits. They're responsible for Cinderella 2 and 3, Bambi 2, Pocohontas 2, Mulan 2, Tarzan 2, The Lion King 1.5 and 2, The Jungle Book 2, Lilo and Stitch 2, and a host of others, most of which can be found wherever used DVDs are sold. So grinding out Star Wars 7 is in line with the established Disney production pipeline.
Yes, Star Wars 7 is nominally a live action film, but today that's just a few principal characters on top of CG animation. Most of the pixels come from the animation teams.
That one word: boring.
Right. Just like commercial air travel, elevators, and escalators. Which is the whole point.
This will be just fine with the trucking industry. The auto industry can deal with "boring" by putting in more cupholders, faux-leather upholstery, and infotainment systems.
after a story I wrote...
This is just self-promotion. Go away.
The US areas that are in trouble are mostly the Gulf coast, especially the Mississippi River flood plain and Florida. Florida is just barely above sea level now, and is very flat.
Slight rises in sea level cause problems all along the Mississippi. Hurricane and storm driven flooding are already getting worse.
The West Coast isn't so bad off, because there are cliffs along most of it. SF, LA and San Diego do have low spots, but they're a few miles long, and seawalls could be built. It might be necessary to dam the SF bay, with something like the Thames Barrage at the Golden Gate.
There's a whole cult of "vertical farming". With LED lighting, it's much more cost-effective. Not only is the power consumption down, but the plants can be racked much more closely without overheating. Phillips has some special-purpose farming LEDs with spectra chosen for growing specific crops.
So far, most of the enthusiasm for this comes from the organic "farmer's market" crowd, not production-scale farmers.
Too many web sites which run ads are buying them through a chain of multiple resellers. Under current law, the web site running the ad can usually disclaim responsibility for hostile ads. That may change. The article is about testimony before the U.S. Senate's committee on homeland security.
The site that displays the ads should be held responsible. Sites which run ads would then need to protect themselves by legal and technical means. For example, if you run ads on your site, your contract with the advertising provider should provide that they will indemify and defend you should a bad ad get through.
The PS3's Cell processor is, in simplified terms, a general purpose CPU and six special purpose coprocessors.
The "coprocessors" are decent general-purpose computers. The problem is that they only have 128K of RAM each. They can access main memory, with high latency but good bandwidth, through a DMA mechanism. (There's also a relatively conventional NVidia GPU on the back end, so the GPU part of the job isn't the big difference. This isn't about shaders.)
That 128K of RAM is too small for a frame. Too small for a level. Too small for any big part of game state. PS3 programming thus consists in turning a problem into some form of streaming process, where data is pumped into Cell processors, processed a bit, and pumped back out. This is a signal processing architecture. It's great for audio. Sucks for everything else.
The PS3 was a "build it and they will come" architecture. It was cheap to make, but large numbers of smart people spent years trying to figure out how to use the thing properly. (Sony basically gutted their US R&D group to beat on that problem.) In practice, a lot of games did most of their work in the main CPU (a MIPS machine) and the GPU, using the Cell processors only for audio, fire and explosions, particle systems, and other tasks that didn't have a lot of interconnected state.
And they've already got it.
No, that's an ad for "Tile". Range 50 to 150 feet to nearest iPhone with the Tile app. Lifespan of one year.
It's another one of those "preorder" scams - pay now, get delivery someday, maybe. "With your help, 49,586 backers preordered Tiles totaling $2,681,297." They're vague about how many were actually delivered.
This is the kind of anti-competitive behavior that gets companies in trouble and causes regulatory crackdowns. Phone companies that make it hard to switch carriers. domain registrars that make it hard to switch registrars, and banks which make it hard to switch banks have all gotten in trouble for this.
If this is just being publicly announced now, it was probably known to NSA, the GRU, and the MSS years ago. The superpower security agencies put substantial resources into cryptanalytic number theory.
Early public key cryptosystems used the knapsack problem. That turned out to have reasonably easy solutions. Factoring products of two big primes may be all we have left. That's believed to be hard, but there's no proof of a lower bound on it.
Modern computers have an IOMMU.
If only. IBM mainframes have had an IOMMU since the IBM 360/67. On IBM mainframes, channels work through the memory protection system, and driver programs can't get a device to read or write outside the program's own space.
PCs and their descendants do not have that. In the early days, MMUs took too many transistors. The Intel line didn't have an CPU-side MMU until the 386 (when 32 bits came in) and Microsoft didn't use the MMU until Windows NT came out in 1993. So there wasn't much call for an IO-side MMU, until it was too late. There are some memory-mapping gimmicks on I/O buses today, but they're mostly there to deal with 32 vs. 64 bit addressing problems.
Peripherals in the PC world have been able to blither all over memory since the earliest days. It's gotten a little better; at least USB devices can't do that. That's important, because, in practice, attacks via USB are common. The buses that offer full memory access are usually internal to the machine, so plugging in a hostile device isn't easy. There are exceptions, including PCMCIA and ExpressCard, which do expose a bus. Exposing a bus on a cable is a very bad idea in today's hostile environment.
Manufacturing auto batteries should be a low-margin business. They're a commodity. Others can enter the business. Over time, margins will decrease. There's not much brand value. (Who made the battery in your phone?)
The article doesn't give the revision history of how that code got there. Who put it there, when did they put it there, and what did the code look like before and after they put it there. We need the names!
Most of the "patent troll" problem comes from three law firms. They're the only ones who've sent out more than two demand letters, according to the EFF.
Read "American Drive", by someone who did a startup in Detroit. His approach was to buy a failing GM axle plant cheap using money from a private equity firm, kick out the union, cut wages over 50%, put in some decent machinery, and make big bucks. That's a Detroit startup for you.
The most amusing part is how they dealt with the crack house across the street from their parking lot. They weren't getting much help from the Detroit cops. So they put stadium-sized lights on the light poles in their parking lot and aimed them all directly at the crack house. When those were switched on for the first time, it was like spraying an ant nest. People ran from the house. The crack house went out of business after a few weeks under the lights.
We need to change the image of "the cloud". Here's the new meme:
The cloud is a cold, dark place. Danger and evil lurk in the cloud, waiting for a chance to strike. You won't see it until it's too late.
The cloud offers temptation. Surrendering to that temptation has a high price. Someday soon you will pay.
Know the cloud. Fear the cloud. Watch your back.
Parking garages, though... Lighted arrows and lines to direct people to empty slots might be useful. It could be useful for intersections which have highly variable traffic patterns, where adding additional turn lanes dynamically is useful. Stadiums often do that, with a small army of people moving traffic cones around.
Solar powered snow melting seems unlikely to work. If you really need snow melting, the power requirements are huge. The cutting edge of technology there is induction heating of snow in railroad switches. Many railroads in snowy areas heat their switches. But nobody heats the entire track.
Ah, startup battery companies. There have been so many.
I have no idea if this will work. But the company seems flaky. Their "path to market" says "This facility will allow Power Japan Plus to meet demand for specialty energy storage markets such as medical devices and satellites." In other words, they're nowhere near making auto batteries.
The medical and satellite industries won't be interested because they already have specialty batteries with demonstrated long life and huge numbers of charge/discharge cycles. Cost isn't a big issue there. So this sounds like an excuse for stringing along investors.
What these guys need to do is build some prototype batteries, put them in a car, and drive on battery power from Tokyo to Osaka without a recharge. That's a little beyond what a Tesla can do. If they can do that, people will listen.
Thunderbolt isn't going to replace USB in all cases, but Thunderbolt isn't about the speed. It's about the protocol. Thunderbolt is basically PCI-E over a wire.
Bad idea for security reasons. Any device plugged in can read and write memory. That's not a good thing. At least with USB, it's just packets to and from the driver.
FireWire had the same problem. Most FireWire PC interfaces allowed limiting the hardware capability to accept packets that read and wrote memory. (There were address limit registers. The default settings for Linux left memory wide open to FireWire attack. (Under Linux, all of memory was open on 32-bit systems, but because this was a bug, not a feature, only the first 4MB was open on 64-bit systems. I once reported this as a Linux kernel bug. There were people who didn't want it fixed because they were using it for kernel debugging.)
Being so close to water and all, the entire idea of an e-reader for naval sailors is preposterous.
This is for submariners. If they encounter water, they have much worse problems than e-reader failure. It's just for recreational reading, anyway; it's not for storing manuals. What the Navy wants is a device which absolutely cannot emit any kind of signal and possibly give away a sub's location.
It might even be waterproof. There are waterproof e-readers. The Navy tends to waterproof everything to at least the "resists high humidity and salt spray" level. That's so routine it's not usually mentioned. (There's no real problem waterproofing electronics. More mobile phones should be waterproof. Some already are; I have one.)
Tell a few of their customers to ship back items which appear to have been tampered with, and compare them at your end. That's appropriate tech support. You have no idea who's doing the tampering or why, and it's worth finding out.
The problems had nothing to do with the use of thorium fuel. It had everything to do with a badly designed cooling system that used He instead of water.
The Ft. St. Vrain story is rather sad. The plant had a large number of minor problems that made it too expensive to run. It was converted to natural gas.
Every reactor design which had something complicated happening within the radioactive parts of the system has been a commercial failure. Standard boiling-water reactors and pressurized-water reactors are very simple both mechanically and chemically inside the reactor vessel. All the complexity is outside, where it can be fixed if necessary.
Sodium-cooled reactors have sodium fires. Pebble-bed reactors have jams. (There's a prototype in Germany that's so jammed it can't be decommissioned.) Helium-cooled reactors have leaks. Reactors which require an adjacent chemical processing plant have all the problems of a chemical plant for radioactive materials. Anything which goes wrong in the radioactive part of the system is a huge deal to fix. The history of exotic reactor designs is not good. Many of the exotic ideas have been funded and built, but the results are not impressive.
Meanwhile, boring old BWR and PWR reactors have a long life and good uptime.
I don't use any Google account services. My mail goes to an IMAP server with spam filtering. The Linux desktop, the Windows 7 desktop, and the Windows laptop all run Thunderbird. The Android smartphone, which does not have a Google account, has an IMAP client. All devices sync mail through IMAP. Works fine.
No ads. Who needs Google?
Multidimensional arrays in C and C++ can be iterated over efficiently along any axis of your choosing. You simply choose which axis when you create the array.
That has to have been written by someone who doesn't do much matrix work. A basic matrix multiply iterates in row order for one matrix and column order for the other.
Easily fixed with libraries like Eigen ( http://eigen.tuxfamily.org/ind... ) and many others.
That's the problem. There's no one way to represent a multidimensional array in C++. There are many ways. Which means math libraries using different ones are incompatible with each other. The last time I did a big number-crunching job in C++, I had four different array representations forced on me by different libraries.
Because the compiler has no clue what those array libraries are doing, you don't get basic loop optimizations that FORTRAN has had for 50 years.