The comment section of a community is a great example of how valuable the community is, and Reddit consistently rewards short dumb posts or jokes, but any sort of debate or intelligent thought contrary to popular opinion is usually down-voted to oblivion.
Their voting system is fundamentally broken, and geared towards flash rather than good content.
Slashdot's voting system -- in comparison -- is great, which is why Slashdot has lasted as long as it has (21 years!) even though it's usually days late to the newest news.
>" Bullshit. GPS will frequently report ridiculous velocities as an inaccurate fix can jump from one side to the other of its approximate location. The further you go (and the longer it takes you), the more accurate the average becomes.
Not only that, but generally the GPS string contains information on the fix quality so you can allow for what uncertainty exists."
That is not true. GPS does not give you velocity, it is extrapolated by the delta in two or more data points over time.
As you said, best case scenario, each data point will have 4m accuracy vertically, horizontally it's closer to 10m. When you account for steeper grades at lower speeds, your accuracy can be way off. If your velocity increases, the delta between the data points are greater, therefor the effects of jitter are less noticable.
All this, add in that those individual data points don't get stored, and usually an average get's stored, you quickly realize that in a best real world case GPS on a phone really can get you down to the level of a city block, that's about it.
Don't beleive me? Take your phone out for a walk, with an app that allows you to export all the GPS data points to a gpx file, and import them to google earth, then, trace the path you actually walked in google earth.
You'll see the GPS path will show multiple changes on all three axis by quite a wide margine.
Now think: People don't hold their phones in the optimal location for GPS. So this will further degrade the data integrity.
Now, I don't know about you, but I've been within a city block of a crime before that I had no involvement in or knowledge of, yet without a doubt, GPS could have places me at the scene of the crime.
But here's the thing, GPS is line of sight, which means likely the location data was based off cell towers.
As you said, the accuracy of this data is measured in kilometres. We've now gone from a city block to 6-10 city blocks along two axis. A single circular KM is a lot of space.
Now moving to accelerometer data, It's meaningless, there is no way to conclusively tell what the person was doing, guess maybe, but prove beyond a reasnoble doubt? No way.
But here's the most daming part of this info, the motion API in iPhones.
What these apps get is an "activity" status, and a "confidence" status. Not raw data.
Location is also garnered via guesses using things like SSID names in surrounding areas correlated with historic GPS data, a method that has to be used since most of the time, the phone is in a pocket and can't get a GPS signal at all.
That kind of data should not be enough to garner a conviction, or even be admissible, it's so horribly flawed.
I suggest you look at the raw data coming out of your phone. It really isn't very accurate at all.
I'm guessing that this bug is in the code morphing software that's on all modern CPUs that intel stole from Transmeta.
As I underetand it, the speed bump in x86 CPUs that's happened in the last decade was by Intel going to RISC processors and running Transmetas code morphing software on their CPUs that dynamically combines multiple x86 CISC operations into fewer, larger RISC ones.
Intel stole the idea from Transmeta back in the mid '00s and got caught and successfully sued for it. They then licenced the technology and used it in every processor since the Core line of processors.
...and one thing that sticks out in memory is in some of their stores, they had opted to build things against code and safety guidelines and just pay the fines as long as it wouldn't shut down the stores.
The company got in trouble regularly because of things like people hurting themselves by walking into glass doors that are hard to see without appropriate markers and whatnot.
The only claim I made was to the difficulty of doing it, not the actual action of doing so.
Fun fact, a lot of keyboards, like Apple keyboards have a few MB of NAND flash for their couple K of unsigned and unencrypted firmware.
Someone made a proof of concept virus that infects those keyboards and uses the flash memory as a keylogger, thereby making the actual keynoard a keylogger.
So this sort of thing has been possible for a while.
Sure the big ISPs can't see what type of data you're sending, but they could check to see if the traffic is routing to ZT's network and just bill it at the highest rate.
I've been praised for being smart my whole life and in the few situations where I was in a position in an exam in which I was privy to some of the answers, I intentionally threw those answers as to not interfere with my test score.
Since they are so insistent on their system being secure when it clearly isn't, wouldn't it be funny if someone sold themselves a ticket with a negative value attached, thereby crediting themselves a large sum of money?
Mine is far less than that for ads. In fact, my attention is almost immediately lost as I hunt through AdBlock settings to figure out how the ad got past my ad blocker.
...5 GHz, which has better range and performance (yes, I know 2.4 has better signal range, but the higher interference leads to a higher usable range in 5GHz in many real-world applications.
No it doesn't. 5GHz has trouble going through simple walls, it's almost line of sight only.
Do a site survey and you'll learn this pretty quick.
In the next 5-10 years, no one is going to have landline internet to their homes anymore.
Mobile data will do to the ethernet jack what cellphones did to the home phone line.
There is no point in redesigning the ethernet jack because in 15 years, no one will be using them anymore. (Shy of in server rooms and data centres, but even then, size won't be the driving force.)
Slashdot Mirror
Very true ^
The comment section of a community is a great example of how valuable the community is, and Reddit consistently rewards short dumb posts or jokes, but any sort of debate or intelligent thought contrary to popular opinion is usually down-voted to oblivion.
Their voting system is fundamentally broken, and geared towards flash rather than good content.
Slashdot's voting system -- in comparison -- is great, which is why Slashdot has lasted as long as it has (21 years!) even though it's usually days late to the newest news.
Actually, if you read their terms of service, if Apple is no longer able to provide this content to you, then you are entitled to a refund.
It was the same deal with the Breaking Bad incident a few years back. They got sued over it, and gave everyone a refund.
>" Bullshit. GPS will frequently report ridiculous velocities as an inaccurate fix can jump from one side to the other of its approximate location. The further you go (and the longer it takes you), the more accurate the average becomes.
Not only that, but generally the GPS string contains information on the fix quality so you can allow for what uncertainty exists."
That is not true. GPS does not give you velocity, it is extrapolated by the delta in two or more data points over time.
As you said, best case scenario, each data point will have 4m accuracy vertically, horizontally it's closer to 10m. When you account for steeper grades at lower speeds, your accuracy can be way off. If your velocity increases, the delta between the data points are greater, therefor the effects of jitter are less noticable.
All this, add in that those individual data points don't get stored, and usually an average get's stored, you quickly realize that in a best real world case GPS on a phone really can get you down to the level of a city block, that's about it.
Don't beleive me? Take your phone out for a walk, with an app that allows you to export all the GPS data points to a gpx file, and import them to google earth, then, trace the path you actually walked in google earth.
You'll see the GPS path will show multiple changes on all three axis by quite a wide margine.
Now think: People don't hold their phones in the optimal location for GPS. So this will further degrade the data integrity.
Now, I don't know about you, but I've been within a city block of a crime before that I had no involvement in or knowledge of, yet without a doubt, GPS could have places me at the scene of the crime.
But here's the thing, GPS is line of sight, which means likely the location data was based off cell towers.
As you said, the accuracy of this data is measured in kilometres. We've now gone from a city block to 6-10 city blocks along two axis. A single circular KM is a lot of space.
Now moving to accelerometer data, It's meaningless, there is no way to conclusively tell what the person was doing, guess maybe, but prove beyond a reasnoble doubt? No way.
But here's the most daming part of this info, the motion API in iPhones.
What these apps get is an "activity" status, and a "confidence" status. Not raw data.
Location is also garnered via guesses using things like SSID names in surrounding areas correlated with historic GPS data, a method that has to be used since most of the time, the phone is in a pocket and can't get a GPS signal at all.
That kind of data should not be enough to garner a conviction, or even be admissible, it's so horribly flawed.
I suggest you look at the raw data coming out of your phone. It really isn't very accurate at all.
Because this data is VERY flawed. GPS has been shown to exaggerate travel distance by as much as 20% on average.
And GPS is super precise by comparison, but relying on cell tower and gyro/accelerometer data?!
Anyone who's tried to correlate this data with any real world measurements knows that the data is about as accurate as throwing darts blindfolded.
I hope this guys lawyer moves to get that evidence thrown out of court.
It definately doesn't belong there.
The article states the flaw dates back a decade, not 20 years.
I'm guessing that this bug is in the code morphing software that's on all modern CPUs that intel stole from Transmeta.
As I underetand it, the speed bump in x86 CPUs that's happened in the last decade was by Intel going to RISC processors and running Transmetas code morphing software on their CPUs that dynamically combines multiple x86 CISC operations into fewer, larger RISC ones.
Intel stole the idea from Transmeta back in the mid '00s and got caught and successfully sued for it. They then licenced the technology and used it in every processor since the Core line of processors.
Or have vision problems... which was the case.
A great thing to not care about for a company that "cares about inclusivity".
...and one thing that sticks out in memory is in some of their stores, they had opted to build things against code and safety guidelines and just pay the fines as long as it wouldn't shut down the stores.
The company got in trouble regularly because of things like people hurting themselves by walking into glass doors that are hard to see without appropriate markers and whatnot.
They didn't give a fuck.
I sure have thought about that, and it would be stupid easy, since DJIs primarlly work connected to an internet connected smart phone.
Speaking more broadly: It's not a question of IF something like this is happening, but rather a question of who's doing it.
The only claim I made was to the difficulty of doing it, not the actual action of doing so.
Fun fact, a lot of keyboards, like Apple keyboards have a few MB of NAND flash for their couple K of unsigned and unencrypted firmware.
Someone made a proof of concept virus that infects those keyboards and uses the flash memory as a keylogger, thereby making the actual keynoard a keylogger.
So this sort of thing has been possible for a while.
Huh, I've always wondered about this.
Everything has powerful CPUs in them now and megabytes of firmware. It wouldn't be hard to do for almost anything.
Add to the fact that most of everything comes from china, manufactured by the lowest bider, it wouldn't be hard.
Sure the big ISPs can't see what type of data you're sending, but they could check to see if the traffic is routing to ZT's network and just bill it at the highest rate.
I've been praised for being smart my whole life and in the few situations where I was in a position in an exam in which I was privy to some of the answers, I intentionally threw those answers as to not interfere with my test score.
I can't say I've ever cheated on a test.
"Linux users celebrated the 26th anniversary by struggling to install nVidia drivers."
Just did a quick google search: https://www.startech.com/ca/Ne...
Apparently 1000BaseT on Coax is still a thing.
There is a lot of bandwidth in those cables...
Like a lot...
But I don't know what else you could use it for other than for SDI.
I'm sure you could use it for data or as a replacement for RJ45, but still... I don't think that would be easy.
Since they are so insistent on their system being secure when it clearly isn't, wouldn't it be funny if someone sold themselves a ticket with a negative value attached, thereby crediting themselves a large sum of money?
Mine is far less than that for ads. In fact, my attention is almost immediately lost as I hunt through AdBlock settings to figure out how the ad got past my ad blocker.
I'm genuinely surprised.
WebKit browsers are so much better. I'm either using Safari or Chrome.
Pros still need a real computer that lets us add and remove PCIe cards.
...like 5 years ago. 10 more dollars and it'll be about what we pay here for activations today.
...5 GHz, which has better range and performance (yes, I know 2.4 has better signal range, but the higher interference leads to a higher usable range in 5GHz in many real-world applications.
No it doesn't. 5GHz has trouble going through simple walls, it's almost line of sight only.
Do a site survey and you'll learn this pretty quick.
Facebook gets preferential treatment. Why shouldn't Spotify?
I think Spotify should sue. I'm sure they will find enough in discovery that would let them win.
In the next 5-10 years, no one is going to have landline internet to their homes anymore.
Mobile data will do to the ethernet jack what cellphones did to the home phone line.
There is no point in redesigning the ethernet jack because in 15 years, no one will be using them anymore. (Shy of in server rooms and data centres, but even then, size won't be the driving force.)