It's custom. I should know, as it's largely designed by me. I worked for Delphi who was the OEM that made the radios for GM. They didn't have any Linux experience, so I was hired for this project. They had lots of talented engineers who experience with VxWorks, QNX, uITRON, etc., but not POSIX/Linux. So I got them up to speed on Linux, helped designed the base OS, and made the Linux system work. I was told the Linux based software generated around $2 billion in sales. Then I got laid off.
It's based on Freescale's LTIB, which I also worked on when I worked for them. But, it's highly customized. Freescale hasn't really maintained LTIB for some time, despite continuing to use it as the base for their BSPs. Something like Debian is much too bloated for what the radios are expected to work it. The same system is used for the simpler non-touch radios in other GM vehicles. It's an ARM9 based iMX25, running at I think 380 MHz, with 32 MB of RAM. The framebuffer comes out of the RAM too. I managed to get it to boot from power on (or rather CAN bus wakeup pulse) through u-boot, through the kernel startup, to system startup and daemons running and have userspace application code start in around 250 ms. Getting the backup camera working in <1 second is an important requirement. The ARM9 as a VIVT cache, which forces a cache flush on each context switch, making it quite slow. If one used udev like a normal Linux distro, it takes something like 3 seconds just for udev to populate/dev on system startup. So obviously udev is out.
The radios are not designed to let you easily root or put different software on. However, stopping someone who has physical access to the radio from hacking it wasn't a very high priority. By default LTIB gives you a blank password root account and a telnet daemon configured to allow root logins! I got rid of that and made it ssh only. I don't know if the final production firmware still has ssh running or not. The iMX53 processor used in the CUE system doesn't have secure boot like some other iMX processors. Freescale's iMX line is actually composed of multiple utterly different families of ARM based SoCs based on different IP. So you can easily hack it with a flash programmer.
I don't know of any easter eggs.... GM isn't the kind of company that would look kindly on that. However, unless someone managed to fix it, you can lockup the DVD player with the "Justice League: Starcrossed" DVD a few minutes in. After the alien ship shoots at some jets. It's not the DVD player, but the video overlay on the iMX53 that has locked up.
Will you still feel the same way when P&G and J&J decide they can compete with you by finding some of their 20 million patents that are possibly related to your product and then keeping you in litigation until your $1m dries up?
I'm using tracfone too, since it's the cheapest I could find for minimal usage. The minutes are pretty cheap ( 10 cents), they don't expire, and the minimum monthly average you need to spend to keep you phone active is really low.
But the SIMs and phones are locked. You can't use a normal SIM in a tracfone or a tracfone SIM in a normal phone. You can't even move tracfone SIMs between tracfones. The phone locks itself to the SIM it's first paired with.
So what if Bob is actually a HFT bot as well? He noticed the pattern of cancelled offers from Eve and let himself get bid up to $24.40 on purpose. After Eve buys from Alice, Bob cancels his order, leaving Eve holding the shares.
Oh yeah, Alice and Bob, they're the same bot.
I'm a developer of Linux based car radios for a major OEM, so I'm getting a kick of of these replies.
Linux can boot fast. A lot has to do with the hardware itself. Managed NAND can take a long time to initialize, SPI NOR flash has a low bandwidth, etc. The manufacturers are asking for things like 1 second boot time. A major concern is how it takes to get the back-up-camera working.
I've managed to get a radio booting from POR to userspace in about 220 ms, and getting to done with Linux system startup by 500 ms. The leaves 500 ms for the application software to startup which was enough to meet the customer's requirements.
Is there a prepaid plan that doesn't require you to pay per month? You buy a set amount of data that you keep until you use it up? I'd like to be able to get data about two weeks a year when I'm traveling.
You've taken into account special relativity, but there is also general relativity to take into account. Gravity slows time. Near the Earth's gravity time is slowed. For GPS satellites, their 3.9 km/sec speed slows their clocks by 7 us/day, but the reduced gravity at 20,000 km speeds their clocks relative the earthbound clocks by 45 us/day. So in that case the effect of gravity is greater than that of speed, and they gain time relative to us.
Voyager 2 is going about 139 us/day slower than us due to it's speed. Once gravity is taken into account, I'd guess the result is somewhere around 82 us/day slower. Of course it's speed and the gravity it's experienced hasn't remained constant.
We got nothing for taking AP classes, so consider yourself lucky. And we didn't have any CS course at all, but I look the CS AB AP test anyway. Don't think I got any college for it, but it probably helped me get out of taking the intro CS courses there and going straight to the more advanced ones.
But it's not a demonstration of functional ASAT capability. The only reason this missle can hit the the satallite is because it's orbit has decayed to the point where it's a week away from crashing to the Earth. It can't hit normal satellites.
Draft a local ordinance with a big fine for driving an oversized vehicle on a road where they are prohibittied.
Prohibit oversized vehicles on said road.
Profit
There are small towns that exist only to serve as speed traps on highways. They incorporate near a highway and lower the speed limit to 25 mph. The only service the town provides is a police force. The only thing the police force does is write speeding tickets. Their only sounce for income is from these speeding tickets. This income is only spent providing huge salaries to the police force.
I can see truck traps being intentionally created. Find a back road that a sat nav sends trucks down. Incorporate there, narrow the road, start fining trucks.
. But we shouldn't even consider building any until we have a *completed* (very) long-term storage/disposal solution for nuclear waste. Deferring it to the next generation is not OK.
What is your (very) long-term storage/disposal solution for waste from coal plants? You know, all that CO2 that's changing our climate. The particulate matter. The sulfur dioxide. The mercury. The heavy metals. The radioactive uranium and thorium. What's your long term plan for all that? Haven't got one?
How about your short term plan? Haven't got one of those either?
Why is it that we must have a (very) long-term plan do deal with storage and disposal of nuclear waste, in some perfectly safe perfectly clean way, when coal plants have no plan what-so-ever to deal with the vastly greater and more damaging waste they generate?
I'm pretty sure the myth of enough plutonium accumulating in some kind of duct to reach criticality, which dispersed it, until it accumulated enough to blue flash again, and so on is "confirmed".
I remember people talking about this back when I worked there, and some of them were actually around when it happened.
unencrypted being the key word there. sci-fi will be encrypted. The only unencrypted signals are the sames ones you can get with an antenna for free, and even then those are encrypted too sometimes.
We'd have no problem with a PCI card that let your turn your computer into a DVR, we just aren't going to manufacture the card. Bullshit! The reason no PCI card with cablecard support exists is because the cable companies won't allow it!. There are companies that would love to sell such a thing, but cablelabs won't allow it. In the rest of the world, it's been possible to buy PCI cards with a CI interface for years. It's only here that the cable companies have prevented any hardware for computers from existing.
And claming that a HD STB costs $1200 is bullshit too. If it was true, the cable would loose money for every customer they signed on. By the time they made back the $1200 (16 years at $6/mo), the STB would need to be replaced.
You cannot use the word "hopefully" that way. Hopefully can only be used as an adverb. A correct usage could be "she waits hopefully." Frankly, Anonymous Coward, I don't give a damn.
In case that went over your head, look up disjunct adverbs.
ivtv is in the kernel now. Some HD cards need firmware and some don't, for example the pcHDTV HD-3000 needs firmware for the Oren demodulator, while the HD-5500 uses a different demodulator and doesn't need firmware.
I've ridden the Shinkansen in Japan. People use them. In Japan we were able to able to meet a friend for breakfast in a city over 150 miles away literally one hour after we walked out the door of our hotel. You can't do anything like that in the United States with our 19th century transportation infastructure. We had trains in this country that ran over 100 mph in the 30s and 40s. Now amtrack's fastest trains go 80 mph and they can't even run them that fast.
In converting from chi-square value to phi correlation you divided out the sample size, which is basically what determines significance. A correlation coefficent alone cannot be used to determine significance. The whole problem with TFA was that he was trying to prove/disprove significance without using an actual significance test. You're right, TFA gets this completely wrong:
If 29% is not considered beyond the reach of chance with respect to 25% in a sample set of 50, what is? While it may initially seem an increase in the sample size might help, if the percentages remain the same, the correlation coefficient does not move.
While it's true that the correlation coefficient doesn't change, the conclusion that increasing the sample size is worthless is wrong. If the sample size increased and the percentages stayed the same, the result would become more significant. Flipping a coin three times and getting heads twice and tails once doesn't mean the coin is unfair. Flipping it 300 times and getting 200 heads and 100 tails does mean the coin is unfair. Different test, but the concept is the same.
I did a little more research, and now I'm pretty sure you're making up that 2.4 second bit. Found this from the insurance insitute for highway safety. Rmember that insurance companies don't make any money from red light fines, but from reducing accidents and thus claims.
Armey's report cites no convincing evidence that yellow signals have been shortened.
Instead, it cites unverified anecdotes -- for example, a television report that the city of
Beaverton, Oregon, shortened the yellow intervals at intersections with cameras "to
accommodate camera enforcement." According to Armey, Beaverton "has been caught
red-handed playing with signal timing on lights that have red light cameras."
The facts indicate otherwise. Armey's office didn't contact local traffic engineers, who
would have explained that signal timing was reviewed when the cameras were installed at
five Beaverton intersections. At two of these, the yellow signal intervals were lengthened,
not shortened, and the intervals at the other three intersections weren't changed (they've
remained unchanged for at least seven years). If the yellow light really is 2.4 seconds, then go video the light to get some real proof and complain to the Oregon State DOT. There are state regulations that municipalities must comply with, and the city of Beaverton can not signal their roads however the hell they want.
There are regulations about yellow time length. The minimum listed in the MUTCD is 3 seconds. I'm going to have to call bullshit on your 2.4 second claim. And what's the speed limit, 25 mph?
Slashdot Mirror
It's custom. I should know, as it's largely designed by me. I worked for Delphi who was the OEM that made the radios for GM. They didn't have any Linux experience, so I was hired for this project. They had lots of talented engineers who experience with VxWorks, QNX, uITRON, etc., but not POSIX/Linux. So I got them up to speed on Linux, helped designed the base OS, and made the Linux system work. I was told the Linux based software generated around $2 billion in sales. Then I got laid off.
/dev on system startup. So obviously udev is out.
It's based on Freescale's LTIB, which I also worked on when I worked for them. But, it's highly customized. Freescale hasn't really maintained LTIB for some time, despite continuing to use it as the base for their BSPs. Something like Debian is much too bloated for what the radios are expected to work it. The same system is used for the simpler non-touch radios in other GM vehicles. It's an ARM9 based iMX25, running at I think 380 MHz, with 32 MB of RAM. The framebuffer comes out of the RAM too. I managed to get it to boot from power on (or rather CAN bus wakeup pulse) through u-boot, through the kernel startup, to system startup and daemons running and have userspace application code start in around 250 ms. Getting the backup camera working in <1 second is an important requirement. The ARM9 as a VIVT cache, which forces a cache flush on each context switch, making it quite slow. If one used udev like a normal Linux distro, it takes something like 3 seconds just for udev to populate
The radios are not designed to let you easily root or put different software on. However, stopping someone who has physical access to the radio from hacking it wasn't a very high priority. By default LTIB gives you a blank password root account and a telnet daemon configured to allow root logins! I got rid of that and made it ssh only. I don't know if the final production firmware still has ssh running or not. The iMX53 processor used in the CUE system doesn't have secure boot like some other iMX processors. Freescale's iMX line is actually composed of multiple utterly different families of ARM based SoCs based on different IP. So you can easily hack it with a flash programmer.
I don't know of any easter eggs.... GM isn't the kind of company that would look kindly on that. However, unless someone managed to fix it, you can lockup the DVD player with the "Justice League: Starcrossed" DVD a few minutes in. After the alien ship shoots at some jets. It's not the DVD player, but the video overlay on the iMX53 that has locked up.
Will you still feel the same way when P&G and J&J decide they can compete with you by finding some of their 20 million patents that are possibly related to your product and then keeping you in litigation until your $1m dries up?
I'm using tracfone too, since it's the cheapest I could find for minimal usage. The minutes are pretty cheap ( 10 cents), they don't expire, and the minimum monthly average you need to spend to keep you phone active is really low. But the SIMs and phones are locked. You can't use a normal SIM in a tracfone or a tracfone SIM in a normal phone. You can't even move tracfone SIMs between tracfones. The phone locks itself to the SIM it's first paired with.
So what if Bob is actually a HFT bot as well? He noticed the pattern of cancelled offers from Eve and let himself get bid up to $24.40 on purpose. After Eve buys from Alice, Bob cancels his order, leaving Eve holding the shares. Oh yeah, Alice and Bob, they're the same bot.
I'm a developer of Linux based car radios for a major OEM, so I'm getting a kick of of these replies. Linux can boot fast. A lot has to do with the hardware itself. Managed NAND can take a long time to initialize, SPI NOR flash has a low bandwidth, etc. The manufacturers are asking for things like 1 second boot time. A major concern is how it takes to get the back-up-camera working. I've managed to get a radio booting from POR to userspace in about 220 ms, and getting to done with Linux system startup by 500 ms. The leaves 500 ms for the application software to startup which was enough to meet the customer's requirements.
Is there a prepaid plan that doesn't require you to pay per month? You buy a set amount of data that you keep until you use it up? I'd like to be able to get data about two weeks a year when I'm traveling.
You've taken into account special relativity, but there is also general relativity to take into account. Gravity slows time. Near the Earth's gravity time is slowed. For GPS satellites, their 3.9 km/sec speed slows their clocks by 7 us/day, but the reduced gravity at 20,000 km speeds their clocks relative the earthbound clocks by 45 us/day. So in that case the effect of gravity is greater than that of speed, and they gain time relative to us. Voyager 2 is going about 139 us/day slower than us due to it's speed. Once gravity is taken into account, I'd guess the result is somewhere around 82 us/day slower. Of course it's speed and the gravity it's experienced hasn't remained constant.
TIn both cases, better training and adhering to safety protocols would have saved lives
Saved lives at Three Miles Island? Who died?
My T61 has a 1394/firewire port. The minidock has a digital coax audio out and DVI, but the laptop itself only has VGA output.
We got nothing for taking AP classes, so consider yourself lucky. And we didn't have any CS course at all, but I look the CS AB AP test anyway. Don't think I got any college for it, but it probably helped me get out of taking the intro CS courses there and going straight to the more advanced ones.
The board gets bonuses too, they all loot the ship together before it sinks.
But it's not a demonstration of functional ASAT capability. The only reason this missle can hit the the satallite is because it's orbit has decayed to the point where it's a week away from crashing to the Earth. It can't hit normal satellites.
- Draft a local ordinance with a big fine for driving an oversized vehicle on a road where they are prohibittied.
- Prohibit oversized vehicles on said road.
- Profit
There are small towns that exist only to serve as speed traps on highways. They incorporate near a highway and lower the speed limit to 25 mph. The only service the town provides is a police force. The only thing the police force does is write speeding tickets. Their only sounce for income is from these speeding tickets. This income is only spent providing huge salaries to the police force.I can see truck traps being intentionally created. Find a back road that a sat nav sends trucks down. Incorporate there, narrow the road, start fining trucks.
What is your (very) long-term storage/disposal solution for waste from coal plants? You know, all that CO2 that's changing our climate. The particulate matter. The sulfur dioxide. The mercury. The heavy metals. The radioactive uranium and thorium. What's your long term plan for all that? Haven't got one?
How about your short term plan? Haven't got one of those either?
Why is it that we must have a (very) long-term plan do deal with storage and disposal of nuclear waste, in some perfectly safe perfectly clean way, when coal plants have no plan what-so-ever to deal with the vastly greater and more damaging waste they generate?
I'm pretty sure the myth of enough plutonium accumulating in some kind of duct to reach criticality, which dispersed it, until it accumulated enough to blue flash again, and so on is "confirmed".
I remember people talking about this back when I worked there, and some of them were actually around when it happened.
unencrypted being the key word there. sci-fi will be encrypted. The only unencrypted signals are the sames ones you can get with an antenna for free, and even then those are encrypted too sometimes.
And claming that a HD STB costs $1200 is bullshit too. If it was true, the cable would loose money for every customer they signed on. By the time they made back the $1200 (16 years at $6/mo), the STB would need to be replaced.
In case that went over your head, look up disjunct adverbs.
ivtv is in the kernel now. Some HD cards need firmware and some don't, for example the pcHDTV HD-3000 needs firmware for the Oren demodulator, while the HD-5500 uses a different demodulator and doesn't need firmware.
It would be less than the airline bailouts were.
I've ridden the Shinkansen in Japan. People use them. In Japan we were able to able to meet a friend for breakfast in a city over 150 miles away literally one hour after we walked out the door of our hotel. You can't do anything like that in the United States with our 19th century transportation infastructure. We had trains in this country that ran over 100 mph in the 30s and 40s. Now amtrack's fastest trains go 80 mph and they can't even run them that fast.
I think I'd rather have the trains.
There are regulations about yellow time length. The minimum listed in the MUTCD is 3 seconds. I'm going to have to call bullshit on your 2.4 second claim. And what's the speed limit, 25 mph?