You could have exposure times longer than 1/framerate. The simplest way would be to use multiple sensors. Say you are shooting 48fps and you want 1/24s exposure. Simply have each of two sensors looking at the same picture recording at 24fps and interlace them into the video. It might look unnatural because of the overlap between each frame, and if people don't like 48fps I doubt they would like this, but besides that I don't see any problem with it.
You might also be able to design a single sensor that gives the same effect, basically by basically accumulating light for the first image from time 0-1/24 and accumulates light for the second image from 1/48-3/48.
I invented a design that sounds like yours. Basically I do this design: http://www.10paperairplanes.com/how-to-make-paper-airplanes/03-the-arrow.html and do a third fold before folding the wings. It makes an extremely fast plane with a deadly sharp point. It won't hold up in a pure distance competition unless you can throw it very fast, but it is accurate and has very low drag.
Don't forget the credit card companies charge merchants fees. Usually something on the order of $0.30 + 2%, but maybe less for a very big merchant.
If pricing is random, as many transactions will round up as round down, meaning the net result is no change. If merchants are careful with their pricing, they might be able to get more transactions to round up, or they can just increase all their prices by $0.025.
That's a fairly standard automated PCB assembly line. I can tell you with 99% certainty that Foxconn workers aren't placing surface mount parts by hand. Machines that can do it are relatively inexpensive (On the order of $1 million) and work many times faster than humans (and many times faster than the machines in that video) with near-perfect accuracy. They will also work 24/7 without creating controversy in western media.
What human workers at Foxconn are doing is final assembly work, such as putting the device together, and maybe packaging it. Robots to do that would be more complicated because a lot of dexterity may be required, and programming them for new products may be difficult because of the wide range of possible movements. For a SMT pick-and-place machine, programming is as easy as giving the machine some design files that tells it where to put the parts, and the basic movements are the same every time. To make a robot that places an assembled board into an enclosure, for example, is more difficult because the enclosure could have any number of shapes.
Here's an example of a modern pick-and-place machine:
http://www.youtube.com/watch?v=nah4BQ9y8IY
Even with a virtually unlimited number of workers, there's no way you could match the output of even one of those machines. Humans make mistakes, so the yield would never be good, and it is extremely difficult for humans to place parts with the fine precision required for ICs with a high pin density.
Ha! 5 years is an eternity for a road around here. There's one section in particular that is resurfaced literally every few months. In just a few months it gets giant potholes that will destroy your car if you hit them. It gets worse faster if it rains (We don't even get much frost here). I don't know why they refuse to dig it up and put in a proper base. Last time I saw them working there, they had removed the road surface to reveal clay soil with big loose rocks (fist-sized, maybe) underneath. It looks like cheap fill that is probably not stable at all especially when wet and probably has no drainage. The base is so rough that when they put down the new layer of asphalt it isn't even smooth (or maybe they just do a bad job with that too).
I can't imagine that getting a crew out there every few months is cheaper than digging down an extra layer and putting down a solid base. They already do 3/4 of the work by tearing up the road and re-paving it, so why not do it right for once?
The Colorado river, which supplies most of the water for southwestern cities like Phoenix and Las Vegas, no longer reaches the ocean. All of the water is diverted and used up along the way, mostly for agriculture.
If you are ever going to be around their warehouse you can place an order and select "will call" for the delivery method and go pick it up yourself a few hours later. The nearest McMaster warehouse is about 30 minutes away, and I've been there a few times. I would love to take a tour, but I'm also happy just to see it from the door.
Their delivery is remarkably fast. One time I placed an order, I don't know what time, but they showed up at my house with it at 8pm the same day in an unmarked van. It was an oversize box, so it was cheaper for them to drive it to me than use UPS.
Georgia has the Hope Scholarship funded by the state lottery. Students who make a 3.0 GPA in high school and maintain that in college will get 90%of tuition per semester paid for, and if they make at least a 3.3 they pay no tuition. Of course, that's not including the $544 and rising "USG Institutional Fee" and various other fees.
Did you work for Inmotion? I have a website there, and I'm wondering if I should move it. This is not the first incident I've had there.
Anyone have a suggestion for a secure and affordable web host?
No, Adblock works. I have it block anything to do with Facebook on third party sites. The like buttons and scripts simply don't load, so Facebook never knows what other sites I go to, and I can leave the Facbook cookie in place.
I haven't been there, but I've been to Kitt Peak Observatory in Arizona, and the one on Mauna Kea (but we couldn't stay on top of Mauna Kea much past sunset, sadly). Mauna Kea was surreal even during the day because it feels like you are on top of the world.
The most amazing view of the night sky I've seen was at Kitt Peak. It really does look a lot like this video, but not as bright and without all the color. To be able to see the sky like this requires an absolutely dark location, and you will be seeing at the lower limit of your vision. The rods in your eyes are more sensitive than cones, but they cannot see color well or at all.
The times in the video when it looks like the sun came up are actually the moon. With a camera, it doesn't take HDR to be able to see stars at the same time as the moon- just set it so the moon is overexposed and the stars are correctly exposed. With your eyes, though, you won't be able to see that many stars once the moon is out because it will temporarily ruin your sensitive night vision. When the sun actually comes up in the video, everything goes white since it looks like they are using a fixed exposure setting.
Last time I tried night sky photography with a camera, I couldn't use long exposures because even in the relatively short exposure time the stars would move enough to leave short trails. They might have done this video by registering several shots from a short time frame on top of each other. Or they might have a better camera than I do. It would take a fast lens and a sensor with low noise at high ISO settings.
In short, it would be worth a trip to camp there, or somewhere similar. You won't see a bright, colorful sky, but you will probably see at least as many stars as there are in this video. Seeing the milky way at night never ceases to amaze me, and it is sad how little of the night sky we can see in most places because of light pollution.
I'm allergic to soybeans, but only mildly. And I'm only allergic to forms containing protein, so soybean oil and soy lecithin are OK (otherwise I wouldn't be able to eat 99% of mass produced food on the market). Soy flour is questionable but OK in small quantities. Hydrolyzed soy protein seems to be OK since it is normally used only in small quantities as a "flavor enhancer", and the process seems to denature the protein enough. Even if I do eat soybeans, I won't die from them since it is a relatively mild allergy. And I'm fairly good at avoiding things with soy.
If I can prove that I'm only allergic to GMO soybeans, I wonder if I could sue Monsanto?
Firewire 800 is slow compared to what modern disks can do, especially SSDs. Here are the approximate sustained speeds I typically get using various interfaces to transfer data to one of my hard drives, a 1TB Samsung F3: USB2: 33MB/sec FW800: 77MB/sec eSATA: 130MB/sec
eSATA is almost 70% faster than FW800 and almost four times as fast as USB2, and I'm still not even using its full bandwidth. Newer SSDs can max out SATA2 at around 250MB/sec, which is why SATA3 was created.
USB2 came out in 2001. Firewire 800 was standardized in 2002. 9 years later, progress should have been made. There's no reason we should have to settle for "fast enough". You might say I don't *need* faster connections to my hard drives, but then I must not *need* anything faster than dial-up and a Pentium 4.
I can here the difference on good headphones and certain songs. I've done listening tests on 3 headphones. Using Sennheiser CX300's ($30), I could barely tell the difference between FLAC and V0 MP3. In a blind test, I might not be able to tell them apart at all. Using Sennheiser IE6's (~$120) it is easier to tell the difference, and I can tell the difference on certain songs. Using my friend's Audio Technica ATH-M50's (~$120), the difference is obvious on many songs. All of this is listening coming straight out of a PC sound card. With the ATH-M50's I will sometimes use a headphone amp because my sound card can't drive them loud enough to hear every detail.
When I say certain songs, I like to pick songs that the V0 encoder had trouble with, or in other words V0 encoded songs with the highest average bitrate. When there are many instruments and voices at the same time, there is more information that has to be represented, or more than may be lost in compression.
The difference is most noticeable in the high end, particularly in the percussion. I can hear up to 19.5KHz. If you can't hear as high, you will have more trouble distinguishing FLAC vs MP3, bu it isn't impossible. Percussion, especially cymbals, sound clearer and sharper in FLAC, but clearness or sharpness alone aren't enough for me to distinguish the difference in back to back listening. I always listen for flaws in the sound of cymbals-- instead of sounding like a real instrument, they will sound digital, shimmery, or wrong. It's hard to describe without making up what sounds like audiophile BS, but the difference is there.
I agree for casual listening FLAC is unnecessary. If there is even the slightest noise coming from the environment (in the car, air conditioning running at home, etc) it becomes hard to distinguish the difference, and if I am not devoting my full attention to the music, I don't hear anything wrong MP3 encoded music.
10G or more can be done on multimode fiber;,that's that 10G Ethernet uses, and I believe up to 25G has been demonstrated over multimode. Some people are also working on 10G or more over plastic fiber.
I did a blind test, and with ATH-M50 headphones, both my friend and I can tell the difference between V0 MP3 and FLAC. I've done the same with my Sennheiser IE6 earbuds, and I can still tell the difference. With my Sennheiser CX-300 earbuds ($30) I can tell the difference if it's not a blind test (though with the right song, a blind test would probably work on those too; I haven't tried it). For casual listening, though, I agree that the differences are too subtle to ever notice or be worth caring about.
One thing that helps is to find a song with a lot of stuff going on all at once (not sure what the technical term for that is). Most of my music s encoded as V0 MP3, and my music program lets me sort by the average bitrate. I have plenty of songs where the average bitrate is 290-310Kbps, which means the V0 encoder needed to use high bitrate in more of the song, and probably hit the bitrate limit (320Kbps) a few times. In other words, these are songs that the encoder has the most trouble with, and thus ones that are easiest to tell apart from FLAC versions. If, on the other hand, you have a song that does not have much going on, like synthesized electronic music with only a few tones, it is easy for the encoder to perfectly represent that in fewer bits.
How much of that silicon is ultra pure semiconductor grade? Probably none, so both materials need to go through a refining process. If there are areas with high moly concentrations, it doesn't matter how much the rest of the world has, as long as those mines are enough to meet demand (and can continue to do so for a while).
I have Clearwire, which is different from Clear. Clearwire is a pre-wimax service sometimes referred to as 3.5G, and our choices for speed are 1.5Mbps or 2Mbps if I remember right. Clear is true Wimax (whether or not you want to call it 4G). Sprint bought Clearwire a while back, and it is unclear what the difference between Clear and Clearwire is. To quote Wikipedia,
Now the company is being marketed under the name CLEAR,[6] except in those markets where the Clearwire name has already been established.[7] (However, it remains uncertain whether this new incarnation of Clearwire, controlled by Sprint, will still continue to offer the contractual conditions which have sparked class action lawsuits in the past.)
I am technically outside their service area in Nevada, but I have a direct line of sight to the tower and I get 5/5 bars of signal on my modem. I normally get about 1.5Mbps down despite being on the 2mbps plan. Apparently my area is oversold. Over Christmas, I started having problems with the internet becoming extremely slow during peak times. It would go from everything being just fine to <80kbps speed and 2-3 second ping times at worst. Once everyone went to bed, it would be fast again. The transition from fast to overloaded would happen suddenly with little warning. Normal ping times to the tower for me are ~60ms. If load increases, ping times may each 200ms but it is still fast enough in terms of download speed to be usable. As load increases, in under a minute it will go from working alright to 1 second+ ping times and super slow download speeds (and stay that way for hours).
Unfortunately Clearwire is really my only option. The neighborhood doesn't have cable, and we are too far from anything to get DSL. There was talk about a company bringing in one or the other (I forgot which), but they would need some large fraction of families in the neighborhood to sign up to make it cost effective for them. The next best option is probably Verizon 3G, but I can only get ~300Kbps down which simply isn't enough (interestingly I get ~700Kbps up though).
They don't have to use refusal as probable cause. There are plenty of signs the officers can look for to see if someone may be drunk- bad driving, glazed eyes, inability to walk in a straight line, smell of alcohol, etc. These methods are often used in addition to a breathalyzer test because the test by itself may not be enough to convict someone (there is some chance the machine was malfuncitoning, for example), but the test combined with conventional signs of drunkenness can prove that someone is drunk beyond reasonable doubt. The same signs of drunkenness can be used as probable cause to get a warrant. If the officer doesn't think you are drunk, he/she will probably just let you go and not ask you to take a test at all.
In many places some kind of test is mandatory. You can refuse the breathalyzer, but instead you have to take a blood test. Probable cause is not involved- it is written into the state laws that if you have a license and drive you consent to either of these tests if asked by the police.
Where I live, there's no DSL or cable, only Clearwire (an early wimax technology inferior to Clear's service. I can't remember if Clear and Clearwire are the same company or not), satellite, or cellular 3G. None of these are particularly good options, but I have Clearwire. It's been OK until recently during peak load times I start getting 1000-2000 ms ping times (around 100 is average) and extremely slow download speeds.
Slashdot Mirror
You could have exposure times longer than 1/framerate. The simplest way would be to use multiple sensors. Say you are shooting 48fps and you want 1/24s exposure. Simply have each of two sensors looking at the same picture recording at 24fps and interlace them into the video. It might look unnatural because of the overlap between each frame, and if people don't like 48fps I doubt they would like this, but besides that I don't see any problem with it. You might also be able to design a single sensor that gives the same effect, basically by basically accumulating light for the first image from time 0-1/24 and accumulates light for the second image from 1/48-3/48.
Georgia Tech published something on this a while back.
http://www.gatech.edu/newsroom/release.html?nid=71506
I invented a design that sounds like yours. Basically I do this design: http://www.10paperairplanes.com/how-to-make-paper-airplanes/03-the-arrow.html and do a third fold before folding the wings. It makes an extremely fast plane with a deadly sharp point. It won't hold up in a pure distance competition unless you can throw it very fast, but it is accurate and has very low drag.
Don't forget the credit card companies charge merchants fees. Usually something on the order of $0.30 + 2%, but maybe less for a very big merchant.
If pricing is random, as many transactions will round up as round down, meaning the net result is no change. If merchants are careful with their pricing, they might be able to get more transactions to round up, or they can just increase all their prices by $0.025.
That's a fairly standard automated PCB assembly line. I can tell you with 99% certainty that Foxconn workers aren't placing surface mount parts by hand. Machines that can do it are relatively inexpensive (On the order of $1 million) and work many times faster than humans (and many times faster than the machines in that video) with near-perfect accuracy. They will also work 24/7 without creating controversy in western media.
What human workers at Foxconn are doing is final assembly work, such as putting the device together, and maybe packaging it. Robots to do that would be more complicated because a lot of dexterity may be required, and programming them for new products may be difficult because of the wide range of possible movements. For a SMT pick-and-place machine, programming is as easy as giving the machine some design files that tells it where to put the parts, and the basic movements are the same every time. To make a robot that places an assembled board into an enclosure, for example, is more difficult because the enclosure could have any number of shapes.
Here's an example of a modern pick-and-place machine: http://www.youtube.com/watch?v=nah4BQ9y8IY Even with a virtually unlimited number of workers, there's no way you could match the output of even one of those machines. Humans make mistakes, so the yield would never be good, and it is extremely difficult for humans to place parts with the fine precision required for ICs with a high pin density.
Ha! 5 years is an eternity for a road around here. There's one section in particular that is resurfaced literally every few months. In just a few months it gets giant potholes that will destroy your car if you hit them. It gets worse faster if it rains (We don't even get much frost here). I don't know why they refuse to dig it up and put in a proper base. Last time I saw them working there, they had removed the road surface to reveal clay soil with big loose rocks (fist-sized, maybe) underneath. It looks like cheap fill that is probably not stable at all especially when wet and probably has no drainage. The base is so rough that when they put down the new layer of asphalt it isn't even smooth (or maybe they just do a bad job with that too).
I can't imagine that getting a crew out there every few months is cheaper than digging down an extra layer and putting down a solid base. They already do 3/4 of the work by tearing up the road and re-paving it, so why not do it right for once?
The Colorado river, which supplies most of the water for southwestern cities like Phoenix and Las Vegas, no longer reaches the ocean. All of the water is diverted and used up along the way, mostly for agriculture.
http://newswatch.nationalgeographic.com/2010/07/29/colorado_river_aspen_environment_forum/
http://green.blogs.nytimes.com/2011/11/17/all-rivers-do-not-run-to-the-sea/
If you are ever going to be around their warehouse you can place an order and select "will call" for the delivery method and go pick it up yourself a few hours later. The nearest McMaster warehouse is about 30 minutes away, and I've been there a few times. I would love to take a tour, but I'm also happy just to see it from the door.
Their delivery is remarkably fast. One time I placed an order, I don't know what time, but they showed up at my house with it at 8pm the same day in an unmarked van. It was an oversize box, so it was cheaper for them to drive it to me than use UPS.
Georgia has the Hope Scholarship funded by the state lottery. Students who make a 3.0 GPA in high school and maintain that in college will get 90%of tuition per semester paid for, and if they make at least a 3.3 they pay no tuition. Of course, that's not including the $544 and rising "USG Institutional Fee" and various other fees.
http://www.finaid.gatech.edu/hopezell/
Did you work for Inmotion? I have a website there, and I'm wondering if I should move it. This is not the first incident I've had there. Anyone have a suggestion for a secure and affordable web host?
No, Adblock works. I have it block anything to do with Facebook on third party sites. The like buttons and scripts simply don't load, so Facebook never knows what other sites I go to, and I can leave the Facbook cookie in place.
Yep, I felt it too. I'm in a 5 story building on the top floor.
I don't know, Monsanto probably has a massive legal team defending them. They are like the RIAA/MPAA of seeds.
I haven't been there, but I've been to Kitt Peak Observatory in Arizona, and the one on Mauna Kea (but we couldn't stay on top of Mauna Kea much past sunset, sadly). Mauna Kea was surreal even during the day because it feels like you are on top of the world.
The most amazing view of the night sky I've seen was at Kitt Peak. It really does look a lot like this video, but not as bright and without all the color. To be able to see the sky like this requires an absolutely dark location, and you will be seeing at the lower limit of your vision. The rods in your eyes are more sensitive than cones, but they cannot see color well or at all.
The times in the video when it looks like the sun came up are actually the moon. With a camera, it doesn't take HDR to be able to see stars at the same time as the moon- just set it so the moon is overexposed and the stars are correctly exposed. With your eyes, though, you won't be able to see that many stars once the moon is out because it will temporarily ruin your sensitive night vision. When the sun actually comes up in the video, everything goes white since it looks like they are using a fixed exposure setting.
Last time I tried night sky photography with a camera, I couldn't use long exposures because even in the relatively short exposure time the stars would move enough to leave short trails. They might have done this video by registering several shots from a short time frame on top of each other. Or they might have a better camera than I do. It would take a fast lens and a sensor with low noise at high ISO settings.
In short, it would be worth a trip to camp there, or somewhere similar. You won't see a bright, colorful sky, but you will probably see at least as many stars as there are in this video. Seeing the milky way at night never ceases to amaze me, and it is sad how little of the night sky we can see in most places because of light pollution.
I'm allergic to soybeans, but only mildly. And I'm only allergic to forms containing protein, so soybean oil and soy lecithin are OK (otherwise I wouldn't be able to eat 99% of mass produced food on the market). Soy flour is questionable but OK in small quantities. Hydrolyzed soy protein seems to be OK since it is normally used only in small quantities as a "flavor enhancer", and the process seems to denature the protein enough. Even if I do eat soybeans, I won't die from them since it is a relatively mild allergy. And I'm fairly good at avoiding things with soy. If I can prove that I'm only allergic to GMO soybeans, I wonder if I could sue Monsanto?
Firewire 800 is slow compared to what modern disks can do, especially SSDs. Here are the approximate sustained speeds I typically get using various interfaces to transfer data to one of my hard drives, a 1TB Samsung F3:
USB2: 33MB/sec
FW800: 77MB/sec
eSATA: 130MB/sec
eSATA is almost 70% faster than FW800 and almost four times as fast as USB2, and I'm still not even using its full bandwidth. Newer SSDs can max out SATA2 at around 250MB/sec, which is why SATA3 was created.
USB2 came out in 2001. Firewire 800 was standardized in 2002. 9 years later, progress should have been made. There's no reason we should have to settle for "fast enough". You might say I don't *need* faster connections to my hard drives, but then I must not *need* anything faster than dial-up and a Pentium 4.
In other words, you could say they ran out of imaginary power.
I can here the difference on good headphones and certain songs. I've done listening tests on 3 headphones. Using Sennheiser CX300's ($30), I could barely tell the difference between FLAC and V0 MP3. In a blind test, I might not be able to tell them apart at all. Using Sennheiser IE6's (~$120) it is easier to tell the difference, and I can tell the difference on certain songs. Using my friend's Audio Technica ATH-M50's (~$120), the difference is obvious on many songs. All of this is listening coming straight out of a PC sound card. With the ATH-M50's I will sometimes use a headphone amp because my sound card can't drive them loud enough to hear every detail.
When I say certain songs, I like to pick songs that the V0 encoder had trouble with, or in other words V0 encoded songs with the highest average bitrate. When there are many instruments and voices at the same time, there is more information that has to be represented, or more than may be lost in compression.
The difference is most noticeable in the high end, particularly in the percussion. I can hear up to 19.5KHz. If you can't hear as high, you will have more trouble distinguishing FLAC vs MP3, bu it isn't impossible. Percussion, especially cymbals, sound clearer and sharper in FLAC, but clearness or sharpness alone aren't enough for me to distinguish the difference in back to back listening. I always listen for flaws in the sound of cymbals-- instead of sounding like a real instrument, they will sound digital, shimmery, or wrong. It's hard to describe without making up what sounds like audiophile BS, but the difference is there.
I agree for casual listening FLAC is unnecessary. If there is even the slightest noise coming from the environment (in the car, air conditioning running at home, etc) it becomes hard to distinguish the difference, and if I am not devoting my full attention to the music, I don't hear anything wrong MP3 encoded music.
10G or more can be done on multimode fiber;,that's that 10G Ethernet uses, and I believe up to 25G has been demonstrated over multimode. Some people are also working on 10G or more over plastic fiber.
I did a blind test, and with ATH-M50 headphones, both my friend and I can tell the difference between V0 MP3 and FLAC. I've done the same with my Sennheiser IE6 earbuds, and I can still tell the difference. With my Sennheiser CX-300 earbuds ($30) I can tell the difference if it's not a blind test (though with the right song, a blind test would probably work on those too; I haven't tried it). For casual listening, though, I agree that the differences are too subtle to ever notice or be worth caring about.
One thing that helps is to find a song with a lot of stuff going on all at once (not sure what the technical term for that is). Most of my music s encoded as V0 MP3, and my music program lets me sort by the average bitrate. I have plenty of songs where the average bitrate is 290-310Kbps, which means the V0 encoder needed to use high bitrate in more of the song, and probably hit the bitrate limit (320Kbps) a few times. In other words, these are songs that the encoder has the most trouble with, and thus ones that are easiest to tell apart from FLAC versions. If, on the other hand, you have a song that does not have much going on, like synthesized electronic music with only a few tones, it is easy for the encoder to perfectly represent that in fewer bits.
How much of that silicon is ultra pure semiconductor grade? Probably none, so both materials need to go through a refining process. If there are areas with high moly concentrations, it doesn't matter how much the rest of the world has, as long as those mines are enough to meet demand (and can continue to do so for a while).
Now the company is being marketed under the name CLEAR,[6] except in those markets where the Clearwire name has already been established.[7] (However, it remains uncertain whether this new incarnation of Clearwire, controlled by Sprint, will still continue to offer the contractual conditions which have sparked class action lawsuits in the past.)
I am technically outside their service area in Nevada, but I have a direct line of sight to the tower and I get 5/5 bars of signal on my modem. I normally get about 1.5Mbps down despite being on the 2mbps plan. Apparently my area is oversold. Over Christmas, I started having problems with the internet becoming extremely slow during peak times. It would go from everything being just fine to <80kbps speed and 2-3 second ping times at worst. Once everyone went to bed, it would be fast again. The transition from fast to overloaded would happen suddenly with little warning. Normal ping times to the tower for me are ~60ms. If load increases, ping times may each 200ms but it is still fast enough in terms of download speed to be usable. As load increases, in under a minute it will go from working alright to 1 second+ ping times and super slow download speeds (and stay that way for hours).
Unfortunately Clearwire is really my only option. The neighborhood doesn't have cable, and we are too far from anything to get DSL. There was talk about a company bringing in one or the other (I forgot which), but they would need some large fraction of families in the neighborhood to sign up to make it cost effective for them. The next best option is probably Verizon 3G, but I can only get ~300Kbps down which simply isn't enough (interestingly I get ~700Kbps up though).
They don't have to use refusal as probable cause. There are plenty of signs the officers can look for to see if someone may be drunk- bad driving, glazed eyes, inability to walk in a straight line, smell of alcohol, etc. These methods are often used in addition to a breathalyzer test because the test by itself may not be enough to convict someone (there is some chance the machine was malfuncitoning, for example), but the test combined with conventional signs of drunkenness can prove that someone is drunk beyond reasonable doubt. The same signs of drunkenness can be used as probable cause to get a warrant. If the officer doesn't think you are drunk, he/she will probably just let you go and not ask you to take a test at all.
In many places some kind of test is mandatory. You can refuse the breathalyzer, but instead you have to take a blood test. Probable cause is not involved- it is written into the state laws that if you have a license and drive you consent to either of these tests if asked by the police.
Where I live, there's no DSL or cable, only Clearwire (an early wimax technology inferior to Clear's service. I can't remember if Clear and Clearwire are the same company or not), satellite, or cellular 3G. None of these are particularly good options, but I have Clearwire. It's been OK until recently during peak load times I start getting 1000-2000 ms ping times (around 100 is average) and extremely slow download speeds.