Apple has released design/block diagrams on the silicon and how "Hey Siri" is implemented in hardware and doesn't require intervention from either the CPU or the OS. It can be verified by putting some scopes and circuit analyzers on the thing and seeing when and where the 'activity' actually happens.
It's fairly easy to test whether or not they're lying, if your CPU and SSD keeps waking up whenever there is audio, even if the trigger hasn't been used, you know they're lying.
Also, you can dump the contents of your iPhone as a developer. So it would also be pretty easy to verify there is no recording lurking somewhere on the drive waiting to be sent to Apple. You could also analyze the traffic that is sent to Apple and see whether it is feasible that audio recordings which would have to be a pretty continuous stream, even encrypted, are being sent without the trigger phrase.
I can get to it since I have subscriptions to these journals:
If you want better P-values, the significance halves. The reporting was done on the raw data which shows a slight variation within the error bars. Even if the study is correct, it comes down that statistically speaking, 2 out of the 500,000 cases may have survived longer if they had a female doctor.
We're talking about a 0.6% difference in mortality rates here on the raw data, well within the error bars. The study itself even shows that for smaller P-values (which basically means increasing statistical accuracy) the difference is about half that or less.
The reporting and abstract painstakingly try to increase the relevance of this study but it seems in ~1,000 cases over all 500,000 was the statistical likelihood of survival larger with a female doctor. This doesn't mean that these people would've guaranteed to live, only that they had a 1 in 60 (or actually closer 1 to 500 chance) of surviving if they had had a female doctor vs a male doctor.
There is a story of this happening in 2014, 2015 and 2017 as well, 2014 is the earliest instance that I could find of that administration lying about net neutrality comments being "hacked" in favor of the consumer.
Bray leaked information to Motherboard in 2014, following another petition to his viewers by John Oliver which caused the website to crash, claiming that malicious activity was responsible.
In this case, it did affect the bank a lot more than the consumer. If you default on your home, you do not keep paying it off. The bank then owns it and has to sell it off for a pittance, often the previous owners will destroy the home as they go out as well and in many cases the houses sit there forever and the bank has to pay the taxes, utilities (in some climates you have to heat the home for it not to freeze), a security company (otherwise looters rip out all the copper) etc on it.
I purchased a house in the middle of the banking crisis a few years ago, banks throw them at your head for about 20% of the evaluated price although they generally need complete renovation. One of the houses I went in had 120yo fine hardwood floors, classic bathrooms, oak doors and trim work, worth about $240k, listed for $80k and every inch of it had deep gauges, all the doors, sinks, bathtubs, stained glass etc were broken.
It's called a mobile home in Europe. I've owned houses that are worth less than a modern RV. Homeless means living on the street or in shelters.
These people are choosing to be "homeless" to save 25k/y in mortgage or rent. When I was young I chose to sleep in my car over renting a hotel. It's a valid choice if you have a temperate climate and a pimped out van, my grandparents lived that way for a while too after retirement in order to travel but they weren't poor, they weren't considered homeless, moreover it wasn't a problem.
This is not a problem that needs fixing, the market will fix itself when people demand to work remotely from a more parochial town where they have a family. Perhaps companies will start spreading (like Amazon has been trying) and setting up in smaller sites in remote places. But as long as people choose to live this way and are happy, I don't see the problem.
Yes: AppleScript (a direct descendant of HyperCard) and LiveCode Dell Boomi HTML w/ JavaScript Old versions of VisualBasic (not the modern VisualStudio)
I see similar applications for very specific purposes. And there is the problem with HyperCard and derivatives: going "beyond" what is natively provided (and thus however much you can possibly fit in the GUI) gets you down a rabbit hole so deep you might as well just learn proper programming.
If you are a 15y sysadmin (25y sysadmin here) then you should know the answers. Containers are an end to a means. You can do yum install and install apache/dns/ldap etc on the same host. Why don't we do that, depends on the use case, in some cases we want to easily spread load or migrate to different physical hosts and the application doesn't have multi-master operations. Full VMs are also a good solution but a bit heavier than containers, if we have hundreds of applications, this matters. Some applications have exploits so you don't want eg. an exploit in DNS that roots the machine affecting your web server.
Yes, I manage 300TB of data for 500 users on ZFS. I know about ACL. I think they're easier to work with than Windows ACL. Snapshots on Windows don't exist, there is a process called Shadow Copy but that's not really what I'd call a snapshot (it takes minutes to complete, relies on the applications to tell the OS that it's okay to be snapshotted, doesn't survive reboots (generally) and rollbacks aren't guaranteed atomic).
I don't think you understand the difference between Windows' restore checkpoints and a true snapshot of a file system. A snapshot restores the data even if the OS doesn't know how to. There are plenty of instances where update rollbacks don't work or make things worse, filesystem snapshots obviate that problem completely by rolling back to a known working state.
Low-frequency 5G (which Verizon and T-Mobile have heavily invested in the past for UHF spectrums) has much larger bandwidth and equal range to 3G and 4G. No way will Verizon set up hundreds of million mini-cells on telephone poles just because. Sure they purchased millimeter wave companies, but that is an entirely different market. 26GHz transmitters in cell phones are going to be bulky and prone to dropouts.
Microsoft should separate out the operating system, the applications, libraries and the user data a lot more and then with proper security measures and filesystem snapshots the operating system could be a lot better, easier to manage and easier to rollback.
If something goes wrong on Linux or BSD I do an apt-get or pkg install with a specific version. With a ZFS boot volume, I snapshot before and after any major updates, something goes wrong it literally takes seconds to repair. On Windows, you can't roll much of anything back without destroying the OS.
Technically, this NES Classic should be able to do the same. It's an ARM system with a stripped down emulator, there are no authentic NES chips in it nor is it a custom chip. It's basically an ODROID2/Raspberry Pi with the Nintendo flavor of Linux.
What you could do though is if you own this, you now own a license to the ROM for use on other emulators.
Word is available on both iOS and Android, tablets aren't intended to be desktop-replacements though. You're not coding or writing papers on them for 8 hours straight, they are simply not dimensioned for that.
There are better Windows tablets than a Surface Pro though. That's the problem, the market is saturated with low-end cheap tablets and iPad-like devices running Linux, Windows and Android (both ARM and x86). The Surface Pro is about twice the cost than its competition.
If you want a productivity device, get a true tablet (Android or iOS), you can get all the "productivity" applications (eg. MS Office, OpenOffice, Terminal emulators, Remote Desktop of various sort (VNC, RDP, X11) and SSH. Windows isn't necessary anymore for anything, except being slow and getting viruses.
I am a sysadmin and as a portable device that can "continue working" during meetings, seminars etc an iPad beats a laptop hands down.
We tried "increasing" fuel economy during the Bush era by giving away money. We ended up with a boom in both SUVs and and an overload of working/newish cars on the scrap heap that we had difficulties recycling. You may not be old enough for the Bush era but government intervention in the market usually has adverse effects.
Most likely the Federal Government or Local Government, they keep track of these things typically in some form of GIS and that is established through documents (eg. deeds). My original deed has the name of some 1800's guy on there that developed the area, so they named it the "developer name" tract and that has changed over time, to a $number ward now and probably many other names (you find things on old maps when you go to city hall for some permit and your name of the area doesn't match their name).
In bigger city, things get renamed whenever a philantropist or developer gets some money in the pockets of the local politicians, in some cases local politicians will change things for like a year or two because of some sort of remembrance or politically correct reason too.
I don't think you fully understand blockchain technology. It's not just about BitCoin (although it's a great example of what kind of trust you can place if used correctly).
Blockchain is nothing but a database "log" basically (if you're familiar what SQL logs are) that is cryptographically chained and verified. I agree that in the logistics world, you still have to trust the sensors (whether human or automatic), but if you can verify that your sensors worked and aren't tampered with, then what's in the blockchain about those sensors is true.
In regards logistics, you can't change the details after the fact to do damage control (eg. glass in the bottles, "oh no, we only keep our logs for a week" will no longer be an excuse) although in the case of logistics, it doesn't solve premeditated negligence (eg. let's disable that glass breakage sensor because it costs us $1000 in profits last week).
A blockchain is a decentralized, untrusted, read-only (as long as one party doesn't own the majority of nodes) database basically.
You can for example, as Nestle, publish the blockchain and then all sorts of consumers and agencies can "join" this blockchain and as long as no single party owns a majority of the nodes, the database can only be written to and never be changed. So you can eg. write all sorts of details about any operation into the database, it gets distributed and then later on, everyone can agree that nobody has tampered with your database.
In the end, in this case, you still have to trust that the translation from physical/environmental variables into electronic input to the chain is not corrupt. If Nestle or their subsidiaries consistently writes "our factory runs the freezer at -10C" into the blockchain but that data isn't true because someone tampered with the sensors, then you still have nothing. But if the blockchain says: the freezer was at +20C for 5 hours at any point in time, the consumer (or anyone in the chain) "knows" something went wrong and you can set up all sorts of searches and alerts that the manufacturer doesn't want you to know.
Slashdot Mirror
We've been herding away birds from fields for several millenia. Should've asked a farmer, not an engineer.
Apple has released design/block diagrams on the silicon and how "Hey Siri" is implemented in hardware and doesn't require intervention from either the CPU or the OS. It can be verified by putting some scopes and circuit analyzers on the thing and seeing when and where the 'activity' actually happens.
It's fairly easy to test whether or not they're lying, if your CPU and SSD keeps waking up whenever there is audio, even if the trigger hasn't been used, you know they're lying.
Also, you can dump the contents of your iPhone as a developer. So it would also be pretty easy to verify there is no recording lurking somewhere on the drive waiting to be sent to Apple. You could also analyze the traffic that is sent to Apple and see whether it is feasible that audio recordings which would have to be a pretty continuous stream, even encrypted, are being sent without the trigger phrase.
I can get to it since I have subscriptions to these journals:
If you want better P-values, the significance halves. The reporting was done on the raw data which shows a slight variation within the error bars. Even if the study is correct, it comes down that statistically speaking, 2 out of the 500,000 cases may have survived longer if they had a female doctor.
We're talking about a 0.6% difference in mortality rates here on the raw data, well within the error bars. The study itself even shows that for smaller P-values (which basically means increasing statistical accuracy) the difference is about half that or less.
The reporting and abstract painstakingly try to increase the relevance of this study but it seems in ~1,000 cases over all 500,000 was the statistical likelihood of survival larger with a female doctor. This doesn't mean that these people would've guaranteed to live, only that they had a 1 in 60 (or actually closer 1 to 500 chance) of surviving if they had had a female doctor vs a male doctor.
There is a story of this happening in 2014, 2015 and 2017 as well, 2014 is the earliest instance that I could find of that administration lying about net neutrality comments being "hacked" in favor of the consumer.
Bray leaked information to Motherboard in 2014, following another petition to his viewers by John Oliver which caused the website to crash, claiming that malicious activity was responsible.
This happened in 2015. Trump got elected 2 years later. Nice try Hillary.
This administration? Wake up, Obama is no longer in office, we elected a golden pumpkin.
Sun Tzu was talking about real human armies. Imagine an army that never tires and the verse no longer applies.
A good compiler for a sane language would simply remove the dead code. If it's not dead code, then it's exploitable.
In this case, it did affect the bank a lot more than the consumer. If you default on your home, you do not keep paying it off. The bank then owns it and has to sell it off for a pittance, often the previous owners will destroy the home as they go out as well and in many cases the houses sit there forever and the bank has to pay the taxes, utilities (in some climates you have to heat the home for it not to freeze), a security company (otherwise looters rip out all the copper) etc on it.
I purchased a house in the middle of the banking crisis a few years ago, banks throw them at your head for about 20% of the evaluated price although they generally need complete renovation. One of the houses I went in had 120yo fine hardwood floors, classic bathrooms, oak doors and trim work, worth about $240k, listed for $80k and every inch of it had deep gauges, all the doors, sinks, bathtubs, stained glass etc were broken.
It's called a mobile home in Europe. I've owned houses that are worth less than a modern RV. Homeless means living on the street or in shelters.
These people are choosing to be "homeless" to save 25k/y in mortgage or rent. When I was young I chose to sleep in my car over renting a hotel. It's a valid choice if you have a temperate climate and a pimped out van, my grandparents lived that way for a while too after retirement in order to travel but they weren't poor, they weren't considered homeless, moreover it wasn't a problem.
This is not a problem that needs fixing, the market will fix itself when people demand to work remotely from a more parochial town where they have a family. Perhaps companies will start spreading (like Amazon has been trying) and setting up in smaller sites in remote places. But as long as people choose to live this way and are happy, I don't see the problem.
Yes:
AppleScript (a direct descendant of HyperCard) and LiveCode
Dell Boomi
HTML w/ JavaScript
Old versions of VisualBasic (not the modern VisualStudio)
I see similar applications for very specific purposes. And there is the problem with HyperCard and derivatives: going "beyond" what is natively provided (and thus however much you can possibly fit in the GUI) gets you down a rabbit hole so deep you might as well just learn proper programming.
If you are a 15y sysadmin (25y sysadmin here) then you should know the answers. Containers are an end to a means. You can do yum install and install apache/dns/ldap etc on the same host. Why don't we do that, depends on the use case, in some cases we want to easily spread load or migrate to different physical hosts and the application doesn't have multi-master operations. Full VMs are also a good solution but a bit heavier than containers, if we have hundreds of applications, this matters. Some applications have exploits so you don't want eg. an exploit in DNS that roots the machine affecting your web server.
Yes, I manage 300TB of data for 500 users on ZFS. I know about ACL. I think they're easier to work with than Windows ACL. Snapshots on Windows don't exist, there is a process called Shadow Copy but that's not really what I'd call a snapshot (it takes minutes to complete, relies on the applications to tell the OS that it's okay to be snapshotted, doesn't survive reboots (generally) and rollbacks aren't guaranteed atomic).
I don't think you understand the difference between Windows' restore checkpoints and a true snapshot of a file system. A snapshot restores the data even if the OS doesn't know how to. There are plenty of instances where update rollbacks don't work or make things worse, filesystem snapshots obviate that problem completely by rolling back to a known working state.
Low-frequency 5G (which Verizon and T-Mobile have heavily invested in the past for UHF spectrums) has much larger bandwidth and equal range to 3G and 4G. No way will Verizon set up hundreds of million mini-cells on telephone poles just because. Sure they purchased millimeter wave companies, but that is an entirely different market. 26GHz transmitters in cell phones are going to be bulky and prone to dropouts.
But... Pai was evil incarnate... net neutrality *twitch* Trump *twitch* Comcast kicked out of NY... WHAT'S HAPPENING!!
Microsoft should separate out the operating system, the applications, libraries and the user data a lot more and then with proper security measures and filesystem snapshots the operating system could be a lot better, easier to manage and easier to rollback.
If something goes wrong on Linux or BSD I do an apt-get or pkg install with a specific version. With a ZFS boot volume, I snapshot before and after any major updates, something goes wrong it literally takes seconds to repair. On Windows, you can't roll much of anything back without destroying the OS.
They do. Lab grown diamonds come in a variety of colors.
Technically, this NES Classic should be able to do the same. It's an ARM system with a stripped down emulator, there are no authentic NES chips in it nor is it a custom chip. It's basically an ODROID2/Raspberry Pi with the Nintendo flavor of Linux.
What you could do though is if you own this, you now own a license to the ROM for use on other emulators.
Word is available on both iOS and Android, tablets aren't intended to be desktop-replacements though. You're not coding or writing papers on them for 8 hours straight, they are simply not dimensioned for that.
There are better Windows tablets than a Surface Pro though. That's the problem, the market is saturated with low-end cheap tablets and iPad-like devices running Linux, Windows and Android (both ARM and x86). The Surface Pro is about twice the cost than its competition.
If you want a productivity device, get a true tablet (Android or iOS), you can get all the "productivity" applications (eg. MS Office, OpenOffice, Terminal emulators, Remote Desktop of various sort (VNC, RDP, X11) and SSH. Windows isn't necessary anymore for anything, except being slow and getting viruses.
I am a sysadmin and as a portable device that can "continue working" during meetings, seminars etc an iPad beats a laptop hands down.
We tried "increasing" fuel economy during the Bush era by giving away money. We ended up with a boom in both SUVs and and an overload of working/newish cars on the scrap heap that we had difficulties recycling. You may not be old enough for the Bush era but government intervention in the market usually has adverse effects.
Most likely the Federal Government or Local Government, they keep track of these things typically in some form of GIS and that is established through documents (eg. deeds). My original deed has the name of some 1800's guy on there that developed the area, so they named it the "developer name" tract and that has changed over time, to a $number ward now and probably many other names (you find things on old maps when you go to city hall for some permit and your name of the area doesn't match their name).
In bigger city, things get renamed whenever a philantropist or developer gets some money in the pockets of the local politicians, in some cases local politicians will change things for like a year or two because of some sort of remembrance or politically correct reason too.
I don't think you fully understand blockchain technology. It's not just about BitCoin (although it's a great example of what kind of trust you can place if used correctly).
Blockchain is nothing but a database "log" basically (if you're familiar what SQL logs are) that is cryptographically chained and verified. I agree that in the logistics world, you still have to trust the sensors (whether human or automatic), but if you can verify that your sensors worked and aren't tampered with, then what's in the blockchain about those sensors is true.
In regards logistics, you can't change the details after the fact to do damage control (eg. glass in the bottles, "oh no, we only keep our logs for a week" will no longer be an excuse) although in the case of logistics, it doesn't solve premeditated negligence (eg. let's disable that glass breakage sensor because it costs us $1000 in profits last week).
A blockchain is a decentralized, untrusted, read-only (as long as one party doesn't own the majority of nodes) database basically.
You can for example, as Nestle, publish the blockchain and then all sorts of consumers and agencies can "join" this blockchain and as long as no single party owns a majority of the nodes, the database can only be written to and never be changed. So you can eg. write all sorts of details about any operation into the database, it gets distributed and then later on, everyone can agree that nobody has tampered with your database.
In the end, in this case, you still have to trust that the translation from physical/environmental variables into electronic input to the chain is not corrupt. If Nestle or their subsidiaries consistently writes "our factory runs the freezer at -10C" into the blockchain but that data isn't true because someone tampered with the sensors, then you still have nothing. But if the blockchain says: the freezer was at +20C for 5 hours at any point in time, the consumer (or anyone in the chain) "knows" something went wrong and you can set up all sorts of searches and alerts that the manufacturer doesn't want you to know.