Doh. Sumbitted parent as HTML not POT. Here's formatting so it's readable.
As someone that works for a major fingerprint sensor manufacturer, I can say that the MythBusters did not select high quality sensors to test against. I'm getting a little tired of people who's entire pool of knowledge about fingerprint sensing is based on this one television making conclusions based on bad information.
I'm not familiar with the door lock sensor specifically, but I can tell from observing it that it is an optical sensor. Whatever live tissue sensing that manufacturer claims to have is obviously not very good because the sensor was defeated with a moist picture of a fingerprint.
The sensor they used on the computer I am familiar with, and it is about three years old. It is, however, based on a much better technology where flesh is live layer of skin is imaged using RF. Unfortunately, the sensor they used is not state of the art, and in fact probably may have been fooled by the circuit board they etched without going to the additional trouble of making the gel finger. Current technology is collecting more information, and is much harder to fool. Which of course implies that it isn't perfect, but it what is?
The real issue for me is "is fingerprint technology increasing or decreasing security?" The narrator made a comment that beating those sensors took them 3 days. In some cases this represents an improvment in security, although not probably for a computer users password (due to brute force password attempts causing lockout..if you have the opportunity to brute force without causing a lock out, then 3 days is probably longer than the password hack would take for most user's passwords.) Think about what the finger print door lock is replacing. One of those mechanical pin locks? Brute forcing a 4 digit pin doesn't take too long, and you can drastically reduce the time required simply by observing someone use the lock and observing a digit or two. You could use a chemical that flouresces under ultraviolet light to see which pins get pressed. For a keyed lock you can just got out and buy a lock picking device. There are certainly cases where even this really bad lock is an improvement.
Regarding the phone, even as strictly a convience feature the fingerprint sensor is a nice to have. The Japanese have been using sensors in phones for years, and they love them. You can think of the sensor as a little touch pad for your phone if you don't want to use it for security. You can cursor around menus and play games using the fingerprint sensor as a touch pad.
Generally, stay away from sensors that only require you to touch the device and hold your finger there. Touch sensors are either optical sensors which need to have some sort of supporing live tissue sensing technology or a really old non-optical sensor. The new non-optical sensors all look like the ones on this phone. A small sliver of silicon over which you drag your finger.
As someone that works for a major fingerprint sensor manufacturer, I can say that the MythBusters did not select high quality sensors to test against. I'm getting a little tired of people who's entire pool of knowledge about fingerprint sensing is based on this one television making conclusions based on bad information.
I'm not familiar with the door lock sensor specifically, but I can tell from observing it that it is an optical sensor. Whatever live tissue sensing that manufacturer claims to have is obviously not very good because the sensor was defeated with a moist picture of a fingerprint.
The sensor they used on the computer I am familiar with, and it is about three years old. It is, however, based on a much better technology where flesh is live layer of skin is imaged using RF. Unfortunately, the sensor they used is not state of the art, and in fact probably may have been fooled by the circuit board they etched without going to the additional trouble of making the gel finger. Current technology is collecting more information, and is much harder to fool. Which of course implies that it isn't perfect, but it what is?
The real issue for me is "is fingerprint technology increasing or decreasing security?" The narrator made a comment that beating those sensors took them 3 days. In some cases this represents an improvment in security, although not probably for a computer users password (due to brute force password attempts causing lockout..if you have the opportunity to brute force without causing a lock out, then 3 days is probably longer than the password hack would take for most user's passwords.) Think about what the finger print door lock is replacing. One of those mechanical pin locks? Brute forcing a 4 digit pin doesn't take too long, and you can drastically reduce the time required simply by observing someone use the lock and observing a digit or two. You could use a chemical that flouresces under ultraviolet light to see which pins get pressed. For a keyed lock you can just got out and buy a lock picking device. There are certainly cases where even this really bad lock is an improvement.
Regarding the phone, even as strictly a convience feature the fingerprint sensor is a nice to have. The Japanese have been using sensors in phones for years, and they love them. You can think of the sensor as a little touch pad for your phone if you don't want to use it for security. You can cursor around menus and play games using the fingerprint sensor as a touch pad.
Generally, stay away from sensors that only require you to touch the device and hold your finger there. Touch sensors are either optical sensors which need to have some sort of supporing live tissue sensing technology or a really old non-optical sensor. The new non-optical sensors all look like the ones on this phone. A small sliver of silicon over which you drag your finger.
The only drivers that must be signed for 32 bit Vista are drivers in the protected media path (DRM) or drivers that load at boot time. Kernel mode drivers for most USB devices, for example, do not require signing on 32 bit Vista.
Components in the Windows Vista Protected Media Path (PMP) must be signed for PMP, and all other kernel-mode components must be signed by WHQL or Kernel Mode Code Signing, in order to ensure access to premium content.
Driver binaries that load at boot time must contain an embedded signature. ...
Slashdot Mirror
He's clearly an engineer. They're always underestimating.
Doh. Sumbitted parent as HTML not POT. Here's formatting so it's readable.
As someone that works for a major fingerprint sensor manufacturer, I can say that the MythBusters did not select high quality sensors to test against. I'm getting a little tired of people who's entire pool of knowledge about fingerprint sensing is based on this one television making conclusions based on bad information.
I'm not familiar with the door lock sensor specifically, but I can tell from observing it that it is an optical sensor. Whatever live tissue sensing that manufacturer claims to have is obviously not very good because the sensor was defeated with a moist picture of a fingerprint.
The sensor they used on the computer I am familiar with, and it is about three years old. It is, however, based on a much better technology where flesh is live layer of skin is imaged using RF. Unfortunately, the sensor they used is not state of the art, and in fact probably may have been fooled by the circuit board they etched without going to the additional trouble of making the gel finger. Current technology is collecting more information, and is much harder to fool. Which of course implies that it isn't perfect, but it what is?
The real issue for me is "is fingerprint technology increasing or decreasing security?" The narrator made a comment that beating those sensors took them 3 days. In some cases this represents an improvment in security, although not probably for a computer users password (due to brute force password attempts causing lockout..if you have the opportunity to brute force without causing a lock out, then 3 days is probably longer than the password hack would take for most user's passwords.) Think about what the finger print door lock is replacing. One of those mechanical pin locks? Brute forcing a 4 digit pin doesn't take too long, and you can drastically reduce the time required simply by observing someone use the lock and observing a digit or two. You could use a chemical that flouresces under ultraviolet light to see which pins get pressed. For a keyed lock you can just got out and buy a lock picking device. There are certainly cases where even this really bad lock is an improvement.
Regarding the phone, even as strictly a convience feature the fingerprint sensor is a nice to have. The Japanese have been using sensors in phones for years, and they love them. You can think of the sensor as a little touch pad for your phone if you don't want to use it for security. You can cursor around menus and play games using the fingerprint sensor as a touch pad.
Generally, stay away from sensors that only require you to touch the device and hold your finger there. Touch sensors are either optical sensors which need to have some sort of supporing live tissue sensing technology or a really old non-optical sensor. The new non-optical sensors all look like the ones on this phone. A small sliver of silicon over which you drag your finger.
As someone that works for a major fingerprint sensor manufacturer, I can say that the MythBusters did not select high quality sensors to test against. I'm getting a little tired of people who's entire pool of knowledge about fingerprint sensing is based on this one television making conclusions based on bad information. I'm not familiar with the door lock sensor specifically, but I can tell from observing it that it is an optical sensor. Whatever live tissue sensing that manufacturer claims to have is obviously not very good because the sensor was defeated with a moist picture of a fingerprint. The sensor they used on the computer I am familiar with, and it is about three years old. It is, however, based on a much better technology where flesh is live layer of skin is imaged using RF. Unfortunately, the sensor they used is not state of the art, and in fact probably may have been fooled by the circuit board they etched without going to the additional trouble of making the gel finger. Current technology is collecting more information, and is much harder to fool. Which of course implies that it isn't perfect, but it what is? The real issue for me is "is fingerprint technology increasing or decreasing security?" The narrator made a comment that beating those sensors took them 3 days. In some cases this represents an improvment in security, although not probably for a computer users password (due to brute force password attempts causing lockout..if you have the opportunity to brute force without causing a lock out, then 3 days is probably longer than the password hack would take for most user's passwords.) Think about what the finger print door lock is replacing. One of those mechanical pin locks? Brute forcing a 4 digit pin doesn't take too long, and you can drastically reduce the time required simply by observing someone use the lock and observing a digit or two. You could use a chemical that flouresces under ultraviolet light to see which pins get pressed. For a keyed lock you can just got out and buy a lock picking device. There are certainly cases where even this really bad lock is an improvement. Regarding the phone, even as strictly a convience feature the fingerprint sensor is a nice to have. The Japanese have been using sensors in phones for years, and they love them. You can think of the sensor as a little touch pad for your phone if you don't want to use it for security. You can cursor around menus and play games using the fingerprint sensor as a touch pad. Generally, stay away from sensors that only require you to touch the device and hold your finger there. Touch sensors are either optical sensors which need to have some sort of supporing live tissue sensing technology or a really old non-optical sensor. The new non-optical sensors all look like the ones on this phone. A small sliver of silicon over which you drag your finger.
The only drivers that must be signed for 32 bit Vista are drivers in the protected media path (DRM) or drivers that load at boot time. Kernel mode drivers for most USB devices, for example, do not require signing on 32 bit Vista.
s ign.mspx
...
...
From http://www.microsoft.com/whdc/winlogo/drvsign/drv
x64 versions of Windows Vista require Kernel Mode Code Signing (KMCS) in order to load kernel-mode software.
Components in the Windows Vista Protected Media Path (PMP) must be signed for PMP, and all other kernel-mode components must be signed by WHQL or Kernel Mode Code Signing, in order to ensure access to premium content.
Driver binaries that load at boot time must contain an embedded signature.