So use a free service like rollernet to operate as a backup MX. If your server goes down, it spools up on their server and then gets pushed through whenever yours comes back online.
Residential customers are not. Industrial customers are. The combustion testing lab where I work has a blowdown wind tunnel. During periods of testing, one of the grad students gets to stay late at night to run the compressor and fill the storage tanks back up at off-peak hours.
And that's the problem... why not make the smart grid simply one with real time pricing? Push the pricing data out of the individual power meters accurate to a minute, let the meters forward that onto the devices that care to know. Have each appliance determine how important it is that it receive that power now, and shut off accordingly. Not going to use your car tomorrow, set it for low priority and it's off if the cost of power comes within 20% of average. Need your car in a few hours, set it in high priority and it won't shut off until power goes 40% above average. Standard deviations might be a better scaling value than percent difference. Add in some fuzzing factor so you don't have a whole block of devices changing state at the same time.
With enough electric cars and other devices operating at night time, power consumption becomes nearly constant. Peaking plants are exchanged for cheaper baseline plants. The real time pricing evolves into less of a means of getting cheap electricity, and more a way for the power company to provide a feedback mechanism to control power consumption. The power company doesn't need any excessively complex system to control individual appliances in each consumer's home, it just needs a feedback system to allow those appliances to regulate themselves.
With the appliances still under customer control, and a bit of randomness thrown into the system, each individual customer's draw will be unpredictable. Combined with a thousand of their closest neighbors tied into the same power plant, the system can be modeled statistically, and a simple tuned PIV loop will alter the price as needed to keep consumption in check.
I think you underestimate just how much oil we use daily. Even if you account for the inefficiencies of internal combustion engines, switching all road vehicles over to electric powered off the grid would nearly double our average power consumption.
That still doesn't explain why with a smart grid, anyone should have any idea of what I am running behind my meter. If I have a trickle charger that draws a modest amount of power 24 hours a day, charging a battery to allow for rapid dumps to my electric car, that would appear to have the same power consumption profile as a grow lamp. That doesn't mean i should tell police ahead of time that is what I'm doing, or have them search my premises to prove my claim. It's none of their business, and I have my right to privacy.
Marijuana growers could burst charge a battery system, and use that to run their grow lamps throughout the day, which would have the same profile as someone charging an electric car straight off the grid. Police are simply going to have to find other evidence to get a warrant.
Yes. It's a closely guarded secret, lest someone else design and construct another tower from scratch, in less time than it takes them to construct, and makes it one foot taller, stealing their thunder. That should eliminate any doubt as to whether this is anything more than a giant architectural penis.
Why would a smart grid let you see what you're running in your house? The most they should have access to are two plots, power and power factor versus time. Anything else is information that neither the utility nor the law needs to know about. There is no need to know what is running inside a home or business, or even have any control over anything running inside it. All the control you need is real time pricing pushed out to the power meter, and from there to connected smart appliances.
Each appliance operates independently, unaware of anything else in the area. Each appliance gets a probability curve where at a certain cost, it may have different behavior. Your car is charging in the garage, and electricity cost is going up. You and your neighbor both have the same charging station running off the same probability algorithm to fit the curve based off daily average power costs. At the current price, your charger has a 50% chance of shutting off and does so. Your neighbor's has a 50% chance, and continues running. Power costs continue to rise, and the chance increases to 60%, at which point his turns off. You're going to be out of town for a week, and simply want your car to be charged when you come back. You tell the unit as such, it automatically adjusts some scaling factors, and now while your neighbor's car is 60% likely to be charging, yours is only 20%. You other neighbor tells his unit damn the cost, never turn off. His unit stays powered up, he pays the incurred cost, and other people's units power down due to increasing price to compensate. Power consuming devices are all turning off in an orderly staggered fashion, with no upstream control.
He means a traditional RPG is story driven, and you progress your character through the story. In Borderlands, the story was ancillary, and served to tie one boss battle to the next. In that respect, it was more like a dungeon grinder (Diablo) or an MMORPG, where the goal is to become more powerful, explore, and collect cool loot, rather than to actually win the game.
It's a valid point, and if you enjoy that style of play, then Borderlands was an amusing game. It's just a terrible way to state it.
Simple reason, it's not a direct competitor. For that quadruple price, you get a 32-bit processor at 5x the clockrate, 150x the storage, and 500x the memory. It's absurdly overkill for the kind of projects people use an Arduino for. It will allow other projects that an Arduino is not capable of. Of course it will be used for simple little things like the Arduino, and in the process lose the greatest utility of the Arduino: teaching people how to write small, efficient, purpose built code.
That's the whole point of the board. It is intended as an Amtel AVR platform for hobbyists to play with. Of course there is no commercial use, because commercial entities would simply fabricate their own system around the AVR microcontroller (and a large number do just this). It's like comparing the BeagleBoard against the use of ARM processors in general.
This product is positioning itself as a microcontroller platform for hobbyists to play with. That puts it firmly in the same market as the Arduino. Again, if commercial users had a need for such a device, they would fabricate their own system based off the ARM7. Now this chip does have the overhead of the.NET runtime environment. On the other hand, it costs 3x as much, has a 32-bit 72MHz ARM, rather than an 8-bit 16MHz AVR, and some 500x the memory. Less of a microcontroller and more of a minicomputer. It's a considerable step up in capability than the Arduino, so while in the same market, it's not a direct competitor.
Who says we will still be using the binary silicon transistor 140 years from now, or even 40 years from now? Who's to say in 15 years when progression of Moore's Law brings us up against uncertainty limits, we won't switch over to some other material, or convert to some other computing paradigm that allows performance to continue its march forward? It's small minded to place any constraints on future technology based on current day knowledge. Similarly, one shouldn't judge future society by today's norms. As western medicine advanced through the 1800s to current, birth rates dropped to prevent the population from growing unsustainably. Power consumption will likely flatten out in a similar manner as we transition over to renewable energy.
So in order to explain why exponential growth of technology is not possible, you gave an example of a technology that has seen fairly consistent exponential growth for the last sixty years? I don't get it.
If Slingbox set up such a system, with one cable box per subscription, and one slingbox per cable box, forwarding the video output to only one endpoint at a time, they would be in violation of terms of service. They would have broken their service contract, and Dish or Comcast would be able to take whatever actions were detailed in the contract. However that is a completely different scenario. In that case, Dish or Comcast is providing a service and can simply rescind the service. In Zevida's case, they have purchased a product, and the movie studio no longer has any say over what happens to that product outside of copyright violation through duplication.
No, not exactly. Their claim to legitimacy was that they purchased dozens of copies of each DVD, and thousands of DVD players paired to slingbox-type products. When the user rented the DVD, the DVD was loaded into a player, plugged into the internet streamer, and unicast to the subscriber. That DVD and DVD player were inaccessible to any other user during that time. It behaved just like a normal rental, except the subscriber did not have to physically go and pick up the DVD. The only real difference between this and a traditional rental service is the turnover rate per disc is much higher.
The Socket 754 boards almost always had three slots, and as you say, the commonly available size for DDR was 1GB, meaning 3GB maximum. The exception was for Mini-ITX systems, and to a lesser extent Micro-ATX, in which board space was at a premium. The Socket 939 boards almost always had four slots, again, for 4GB maximum. I have two of these systems personally, and another 250 on shelves at work. When I purchased mine ~5yrs ago, it was rare to find one with only two slots. You either had to get a small form factor board, or go for really cheap components, which were themselves rare since 939 was marketed to the enthusiast and not the bargain buyer. Of course my experience is from building my own hardware, and someone buying a pre-built OEM system may find two slots to be more common.
However this is not a motherboard limitation, it is a CPU limitation. The CPU drives the memory, and the motherboard just provides the wiring to do so. The 754 chips were only capable of pushing 12 ranks out of its one controller. The 939 chips were only capable of pushing 8 ranks out of each of their two controllers. The motherboards couldn't add another pair of slots even if they wanted to, as the CPU didn't have the power to drive them. The whole purpose of buffered and fully buffered server memory is to reduce the load on the memory controller, and allow it to push three or more slots each.
No. DGPS uses a secondary signal to broadcast an offset, valid for an area a few tens of miles, with diminishing accuracy outside that. RTK combines DGPS with other trickery such as carrier phase enhancement, and a big chunk of extra processing power, to come up with a more accurate position fix.
What you are referring to is the practice of using two sufficiently different carrier frequencies from a single satellite to calculate the ionospheric delay and improve accuracy with a single receiver. This is rarely found outside of military gear. In contrast, DGPS was a civilian enhancement intended to get around the navigational problems caused by selective availability.
I'm not sure I understand how a motherboard could limit the amount of installed memory when the memory controller was built into the CPU. Even still, the original AMD64 chips had no trouble supporting 4 and 8GB of memory. The problem was the memory controller on the original AMD64 chips had to downclock when too many ranks were installed, and it was simply difficult to find large enough DIMMs to actually hit 8GB.
In any case, I don't agree with you on most users not leaving their systems on 24/7. At least in my experience, once people started moving to 'always on internet', they thought that meant they were supposed to leave their PC on all the time too. Either that, or they didn't know about sleep mode, and simply found a minute or so booting up the machine to use it inconvenient.
No. Differential GPS uses a local transmitter to provide local offset to the GPS position. This uses local transmitters as the time source, to allow it to operate in places the GPS signals will not reach.
There are several solutions available currently that offer 1" or better absolute accuracy off of GPS signals. They all function by resolving the issue that GPS is precise, but not accurate. By positioning themselves at a known position in space, they calculate the offset from what GPS is telling them, and broadcast that to the nearby area. They end up being much cheaper as all you need is a good GPS receiver.
This solution instead requires accurate local time references, and as such is going to be considerably more expensive. The advantage of this system is that GPS has an incredibly weak signal. It requires line-of-sight, and even trees will block a signal. If inside a building, or down in a quarry, you are likely to not receive enough signals to get an accurate position lock. This uses local transmitters at high power to allow them to operate in less advantageous terrain.
Good to know. I tried it a long time back for reasons I don't recall, and gave up on it after a several GB, several hour download finished, only for the installer to tell me my processor didn't support it.
No. It takes hardware accelerated virtualization. VirtualPC refuses to run without it. Even in the current lineup, Intel still offers several budget processors with it disabled.
Slashdot Mirror
So use a free service like rollernet to operate as a backup MX. If your server goes down, it spools up on their server and then gets pushed through whenever yours comes back online.
The problem with CRM114 is that it let's anything past if you just put 'POE' in the title.
To hell with satirizing the police, he should be sent to jail for subjecting the public to more inane xtranormal cartoons.
Residential customers are not. Industrial customers are. The combustion testing lab where I work has a blowdown wind tunnel. During periods of testing, one of the grad students gets to stay late at night to run the compressor and fill the storage tanks back up at off-peak hours.
And that's the problem... why not make the smart grid simply one with real time pricing? Push the pricing data out of the individual power meters accurate to a minute, let the meters forward that onto the devices that care to know. Have each appliance determine how important it is that it receive that power now, and shut off accordingly. Not going to use your car tomorrow, set it for low priority and it's off if the cost of power comes within 20% of average. Need your car in a few hours, set it in high priority and it won't shut off until power goes 40% above average. Standard deviations might be a better scaling value than percent difference. Add in some fuzzing factor so you don't have a whole block of devices changing state at the same time.
With enough electric cars and other devices operating at night time, power consumption becomes nearly constant. Peaking plants are exchanged for cheaper baseline plants. The real time pricing evolves into less of a means of getting cheap electricity, and more a way for the power company to provide a feedback mechanism to control power consumption. The power company doesn't need any excessively complex system to control individual appliances in each consumer's home, it just needs a feedback system to allow those appliances to regulate themselves.
With the appliances still under customer control, and a bit of randomness thrown into the system, each individual customer's draw will be unpredictable. Combined with a thousand of their closest neighbors tied into the same power plant, the system can be modeled statistically, and a simple tuned PIV loop will alter the price as needed to keep consumption in check.
I think you underestimate just how much oil we use daily. Even if you account for the inefficiencies of internal combustion engines, switching all road vehicles over to electric powered off the grid would nearly double our average power consumption.
That still doesn't explain why with a smart grid, anyone should have any idea of what I am running behind my meter. If I have a trickle charger that draws a modest amount of power 24 hours a day, charging a battery to allow for rapid dumps to my electric car, that would appear to have the same power consumption profile as a grow lamp. That doesn't mean i should tell police ahead of time that is what I'm doing, or have them search my premises to prove my claim. It's none of their business, and I have my right to privacy.
Marijuana growers could burst charge a battery system, and use that to run their grow lamps throughout the day, which would have the same profile as someone charging an electric car straight off the grid. Police are simply going to have to find other evidence to get a warrant.
Yes. It's a closely guarded secret, lest someone else design and construct another tower from scratch, in less time than it takes them to construct, and makes it one foot taller, stealing their thunder. That should eliminate any doubt as to whether this is anything more than a giant architectural penis.
When I first read that, I thought you said 'tight sweater', and were subtly approving the behavior.
Why would a smart grid let you see what you're running in your house? The most they should have access to are two plots, power and power factor versus time. Anything else is information that neither the utility nor the law needs to know about. There is no need to know what is running inside a home or business, or even have any control over anything running inside it. All the control you need is real time pricing pushed out to the power meter, and from there to connected smart appliances.
Each appliance operates independently, unaware of anything else in the area. Each appliance gets a probability curve where at a certain cost, it may have different behavior. Your car is charging in the garage, and electricity cost is going up. You and your neighbor both have the same charging station running off the same probability algorithm to fit the curve based off daily average power costs. At the current price, your charger has a 50% chance of shutting off and does so. Your neighbor's has a 50% chance, and continues running. Power costs continue to rise, and the chance increases to 60%, at which point his turns off. You're going to be out of town for a week, and simply want your car to be charged when you come back. You tell the unit as such, it automatically adjusts some scaling factors, and now while your neighbor's car is 60% likely to be charging, yours is only 20%. You other neighbor tells his unit damn the cost, never turn off. His unit stays powered up, he pays the incurred cost, and other people's units power down due to increasing price to compensate. Power consuming devices are all turning off in an orderly staggered fashion, with no upstream control.
He means a traditional RPG is story driven, and you progress your character through the story. In Borderlands, the story was ancillary, and served to tie one boss battle to the next. In that respect, it was more like a dungeon grinder (Diablo) or an MMORPG, where the goal is to become more powerful, explore, and collect cool loot, rather than to actually win the game.
It's a valid point, and if you enjoy that style of play, then Borderlands was an amusing game. It's just a terrible way to state it.
Simple reason, it's not a direct competitor. For that quadruple price, you get a 32-bit processor at 5x the clockrate, 150x the storage, and 500x the memory. It's absurdly overkill for the kind of projects people use an Arduino for. It will allow other projects that an Arduino is not capable of. Of course it will be used for simple little things like the Arduino, and in the process lose the greatest utility of the Arduino: teaching people how to write small, efficient, purpose built code.
That's the whole point of the board. It is intended as an Amtel AVR platform for hobbyists to play with. Of course there is no commercial use, because commercial entities would simply fabricate their own system around the AVR microcontroller (and a large number do just this). It's like comparing the BeagleBoard against the use of ARM processors in general.
This product is positioning itself as a microcontroller platform for hobbyists to play with. That puts it firmly in the same market as the Arduino. Again, if commercial users had a need for such a device, they would fabricate their own system based off the ARM7. Now this chip does have the overhead of the .NET runtime environment. On the other hand, it costs 3x as much, has a 32-bit 72MHz ARM, rather than an 8-bit 16MHz AVR, and some 500x the memory. Less of a microcontroller and more of a minicomputer. It's a considerable step up in capability than the Arduino, so while in the same market, it's not a direct competitor.
Who says we will still be using the binary silicon transistor 140 years from now, or even 40 years from now? Who's to say in 15 years when progression of Moore's Law brings us up against uncertainty limits, we won't switch over to some other material, or convert to some other computing paradigm that allows performance to continue its march forward? It's small minded to place any constraints on future technology based on current day knowledge. Similarly, one shouldn't judge future society by today's norms. As western medicine advanced through the 1800s to current, birth rates dropped to prevent the population from growing unsustainably. Power consumption will likely flatten out in a similar manner as we transition over to renewable energy.
So in order to explain why exponential growth of technology is not possible, you gave an example of a technology that has seen fairly consistent exponential growth for the last sixty years? I don't get it.
Sounds like the Dark Ages all over again.
If Slingbox set up such a system, with one cable box per subscription, and one slingbox per cable box, forwarding the video output to only one endpoint at a time, they would be in violation of terms of service. They would have broken their service contract, and Dish or Comcast would be able to take whatever actions were detailed in the contract. However that is a completely different scenario. In that case, Dish or Comcast is providing a service and can simply rescind the service. In Zevida's case, they have purchased a product, and the movie studio no longer has any say over what happens to that product outside of copyright violation through duplication.
No, not exactly. Their claim to legitimacy was that they purchased dozens of copies of each DVD, and thousands of DVD players paired to slingbox-type products. When the user rented the DVD, the DVD was loaded into a player, plugged into the internet streamer, and unicast to the subscriber. That DVD and DVD player were inaccessible to any other user during that time. It behaved just like a normal rental, except the subscriber did not have to physically go and pick up the DVD. The only real difference between this and a traditional rental service is the turnover rate per disc is much higher.
The Socket 754 boards almost always had three slots, and as you say, the commonly available size for DDR was 1GB, meaning 3GB maximum. The exception was for Mini-ITX systems, and to a lesser extent Micro-ATX, in which board space was at a premium. The Socket 939 boards almost always had four slots, again, for 4GB maximum. I have two of these systems personally, and another 250 on shelves at work. When I purchased mine ~5yrs ago, it was rare to find one with only two slots. You either had to get a small form factor board, or go for really cheap components, which were themselves rare since 939 was marketed to the enthusiast and not the bargain buyer. Of course my experience is from building my own hardware, and someone buying a pre-built OEM system may find two slots to be more common.
However this is not a motherboard limitation, it is a CPU limitation. The CPU drives the memory, and the motherboard just provides the wiring to do so. The 754 chips were only capable of pushing 12 ranks out of its one controller. The 939 chips were only capable of pushing 8 ranks out of each of their two controllers. The motherboards couldn't add another pair of slots even if they wanted to, as the CPU didn't have the power to drive them. The whole purpose of buffered and fully buffered server memory is to reduce the load on the memory controller, and allow it to push three or more slots each.
No. DGPS uses a secondary signal to broadcast an offset, valid for an area a few tens of miles, with diminishing accuracy outside that. RTK combines DGPS with other trickery such as carrier phase enhancement, and a big chunk of extra processing power, to come up with a more accurate position fix.
What you are referring to is the practice of using two sufficiently different carrier frequencies from a single satellite to calculate the ionospheric delay and improve accuracy with a single receiver. This is rarely found outside of military gear. In contrast, DGPS was a civilian enhancement intended to get around the navigational problems caused by selective availability.
I'm not sure I understand how a motherboard could limit the amount of installed memory when the memory controller was built into the CPU. Even still, the original AMD64 chips had no trouble supporting 4 and 8GB of memory. The problem was the memory controller on the original AMD64 chips had to downclock when too many ranks were installed, and it was simply difficult to find large enough DIMMs to actually hit 8GB.
In any case, I don't agree with you on most users not leaving their systems on 24/7. At least in my experience, once people started moving to 'always on internet', they thought that meant they were supposed to leave their PC on all the time too. Either that, or they didn't know about sleep mode, and simply found a minute or so booting up the machine to use it inconvenient.
No. Differential GPS uses a local transmitter to provide local offset to the GPS position. This uses local transmitters as the time source, to allow it to operate in places the GPS signals will not reach.
There are several solutions available currently that offer 1" or better absolute accuracy off of GPS signals. They all function by resolving the issue that GPS is precise, but not accurate. By positioning themselves at a known position in space, they calculate the offset from what GPS is telling them, and broadcast that to the nearby area. They end up being much cheaper as all you need is a good GPS receiver.
This solution instead requires accurate local time references, and as such is going to be considerably more expensive. The advantage of this system is that GPS has an incredibly weak signal. It requires line-of-sight, and even trees will block a signal. If inside a building, or down in a quarry, you are likely to not receive enough signals to get an accurate position lock. This uses local transmitters at high power to allow them to operate in less advantageous terrain.
Good to know. I tried it a long time back for reasons I don't recall, and gave up on it after a several GB, several hour download finished, only for the installer to tell me my processor didn't support it.
No. It takes hardware accelerated virtualization. VirtualPC refuses to run without it. Even in the current lineup, Intel still offers several budget processors with it disabled.