The law is typically written to specifically exempt law enforcement and emergency responders.
Most of those exemptions are only truly valid while on an "emergency" response, even though most cops treat them like full-time exemptions.
Because of the emergency requirement, and because the police car has a radio and most have computers, I can't think of any reason to allow cops to use hand-held cell phones for any reason while driving. If they are responding to an emergency, they should be using official communications methods, and if they aren't, they shouldn't get an exemption from laws.
Why? Really why should I wait? I have an app on my phone that reads my text message if I am in my car. I never respond and it is no more distracting than the radio.
Similarly, my car provides this same feature, and also allows me to send "canned" texts (similar to one-button taunts in online games) with a couple of clicks on the car's touch screen. I also can't edit "regular" texts while the car is in motion, but I can edit either the canned texts to meet my needs or edit a normal text while the car is stopped but still hit "send" while the car is in motion.
And if you want to put a 1TB RAM drive into a server?
Buy hardware that supports 2-4TB on the motherboard.
Regardless of the physical size of the server, the price for such a motherboard plus RAM is going to be less than the add-in card plus RAM. It might be a lot more for the add-in solution, as you might not get warranty support unless you buy RAM from the same company that sells the card. You might also have to decide how much RAM disk you want at time of purchase, while system RAM can be upgraded (to the limit of the motherboard) and you can easily decide the size of RAM disk via software.
Also, for Linux, add-in cards might not have drivers, while system RAM does.
Your absolutely right. A lot of iPhone users only want an iPhone due to the traditionally tiny form factor. It is somewhat surprising to see Apple alienate this crowd.
There is no reason the iPhone 6 needs to be larger than the iPhone 5, even with a larger screen.
It's just Apple sticking to the exact same space-wasting layout of previous iPhones that makes it so. My phone has a 5.2" 1920x1080 screen, giving it a bigger screen and more resolution than the iPhone 6, yet the iPhone 6 is only 0.25" less wide (and the same height).
The iPhone 6 is 4.7", and might fit your hand better than the 4.7" phone you had.
The iPhone 5 has a 5.3" diagonal case with a 4" screen. My LG G2 has a 5.2" diagonal screen with a 5.75" diagonal case.
So, if they lay out the iPhone 6 screen so it uses the phone area better, it should be about a 5.2" screen, which is less than the old iPhones. Even the 6 Plus shouldn't be more than 6" diagonal case, if designed correctly.
When my phone rings, just pulling it out of my pocket to check who's calling is actually kind of a pain in the ass--depending on temp and what gloves I'm wearing, sneaking a peek at my wrist is potentially much easier
A bluetooth headset that speaks caller ID is even easier.
Seriously, if you feel you must have your phone with you while actually on the slope, then you don't understand the concept of "time off". If your clients feel they are so important that not getting back to them in an hour or two will cost you their business, then you are even worse off than someone living paycheck to paycheck.
Can you list what dangerous services are turned on by default on a Windows Server install?
"Server" is set to automatic and shouldn't be required at all unless you want to share files to another machine, and is the historic point of access for most exploits.
"Remote Desktop Services" is set to manual but seems to get started by some other service in a default install.
"WebClient" is set to manual but seems to get started by some other service in a default install, and is dangerous because it allows any program to upload/download files via a "trusted" mechanism, even if the program itself isn't really trustworthy.
I'm sure there are a lot more, but all of my machines have had useless startup services disabled, so I'd have to do a completely fresh install to check for sure.
Prime Video really isn't a good comparison to any other service, because it's really just a free extra, since Prime is mostly about free two-day shipping.
I watch Prime videos for exactly that reason...they are free. The $100 annual fee for Prime is well worth it for my household, as we average one package per week from Amazon.
For example, getting sendmail to not start until the clamd server is ready to accept connections isn't easy using systemd, but trivial using a standard init script.
In sendmail.service:
After=clamd.service
This still doesn't fix the problem I had, which was that the clamd socket wasn't accepting connections when sendmail actually started, so sendmail failed.
Also, you can't edit sendmail.service and have the changes stick, as it is located at/usr/lib/systemd/system/sendmail.service and will be overwritten the next time sendmail is updated. I know there is a poorly documented mechanism for overriding the built-in version of a *.service file, but could never figure out how it really works. OTOH, editing/etc/init.d/sendmail works really well and isn't overwritten by a package update, because packagers know files in/etc/init.d are often changed.
A reasonable dependency system would define exactly what "available" means and have the manager (systemd) run code from the definition that is a test for the required condition.
An example would be to connect to a MySQL database and see if you get a response. A better test would be a user-defined query that should return a known value.
Maybe you haven't heard of the X Window System, introduced over 25 years ago. It allows you to run graphical applications remotely, so the system with the application does not need a GUI at all.
Except that the system with the application still needs 273* extra packages installed to allow it to run X client applications.
That's 273 extra code bases that need to be secured to keep the server safe.
* Although not the actual number, it's not far from the truth.
Chip + PIN effectively mitigates the weakness in magnetic strip data by embedding a chip (physical, something you have) and a pin (something you know) into the transaction process, plus many other security enhancements.
Since some of the cards stolen were debit cards, which require something you have (card with magnetic strip) and something you know (PIN), I don't see how chip+PIN is the holy grail you think it is.
Although there may be more negotiation/handshake at PoS with chip+PIN, it still comes down to two-factor auth to make that sale. And, if somebody can install software/hardware that grabbed mag strip + PIN, they likely can do the same for chip+PIN.
Almost everyone I've asked that has expressed hatred of SystemD hasn't actually used it.
I've used it. I hate it.
Ignoring the very real problem that putting so damn much in PID 1 is dangerous for system stability and security, systemd is generally OK for all distribution-supplied packages. But, if you have anything at all that the packagers didn't think of, it's a pain in the ass. For example, getting sendmail to not start until the clamd server is ready to accept connections isn't easy using systemd, but trivial using a standard init script.
Also, despite the fact that dependencies are baked-in to systemd, it's not at all uncommon for a service that depends on an something else (service, NFS mount, etc.) to still start up before the dependency is fully ready, simply because the default systemd is to assume the dependency is fulfilled as soon as whatever "starts" it returns.
Next, there is no easy way to copy existing dependencies to another service (which would be the best way to start creating your own), mostly because the systemd docs and examples simply suck.
Last, the dependency system absolutely screams for a GUI interface to be able to follow and configure it, but when one finally is created (if it hasn't been already), it'll be useless on servers, because nobody with brains installs a GUI on the server.
The US asked an Irish court to issue a warrant to force production of the data. The Irish court refused to issue the warrant.
And the question here is: Why?
I don't know the laws of Ireland enough to know if the reason the judge refused to issue a warrant is part of the public record. I don't even know if the judge has to articulate a specific reason.
And, although it might be interesting to know, it really doesn't matter. All that matter is that a court with jurisdiction to issue a warrant said "no", and the US tried an end-run around that authority.
But in the end, many companies will choose to foot a larger bill if they can budget for it in advance. "We will need $100,000 for Office licenses" often sounds better than "We didn't budget for it, but it turns out we need $30,000 for various unforeseen development expenses."
It also rarely matters what the total cost turns out to be, but rather the per-seat cost.
If you can get a big enough volume discount, $200 or so per seat doesn't look bad considering that it's a purchase (Office doesn't require yearly fees), and you get a lot more support overall (both official through MS and various help web sites).
Some are freaking-out though, or at least their publicity management people are. From what I gather, someone started tweeting they were fake, and apparently the more were released to prove they weren't.
If all the leaked photos are very "private", then I suspect they are fakes.
On the other hand, if there are lots of "boring" pictures, too, it's likely they are real.
Buy a Winegard 8800 and a Boost XT preamp, and that distance should be no problem.
I have an 8800 (and a 4400) and a low noise pre-amp on each. I didn't say OTA TV was unwatchable...it's actually quite good. Some signals are so strong that the antenna pointed 90 degrees from the source still gets a signal well over 24dB C/N. On the "proper" antenna, signals can hit 30dB C/N.
But, DirecTV still has fewer dropouts than even the strongest channels, since multiple reflected signals can wreak havoc, and stronger signals are often affected even more. I can't control airplanes, moving trees, or other intermittent reflectors, and these happen far more often than really thick clouds, so OTA fares worse than satellite.
But if it was an American, residing on European soil, there would be extradition procedures to follow. And those would involve having the local (EU) police generate their own warrant and make their own arrest based upon a formal request.
If you had followed this case, you would know that this is exactly what the US tried to do.
The US asked an Irish court to issue a warrant to force production of the data. The Irish court refused to issue the warrant. So, the US issued a subpoena to Microsoft, who rightly told the US that although the data was on a Microsoft computer, the data was owned by a customer of Microsoft, therefore a warrant would be required. The US court then issued a warrant for Microsoft to produce the data. Microsoft refused, noting that the data was in a foreign country, and warrants are only valid when issued by a court that has jurisdiction over the location of the requested object/data/person. No US court has jurisdiction over Irish soil, thus we end up at today's story.
The actual point of Microsoft's appeal is that the US wants to have a court to be able to issue an order that has the all the advantages of both a warrant and a subpoena, while ignoring their limitatations. The problem with this is that subpoenas are allowed to be fairly vague and apply to anything that is "owned" by the target of the subpoena, regardless of where it is located. Warrants, OTOH, can force the target to hand over something they don't own but over which they have control, but can only request very specific items/data, and have to be issued by a court that has jurisdiction over where the item/data is located.
A short description of what the patch actually is intended to do would not kill Microsoft. I shouldn't have to go hunting for that information if I want it.
In addition, if you have set Windows Update to "download but not install", then it is possible that you don't have Internet access at the time you are thinking of applying the already-downloaded patch.
And LTO is far more reliable than a SATA hard disk.
It depends on what you mean by "reliable". Able to withstand impact, yes. Able to keep bits intact when stored as they should be...both LTO6 and disk are about the same.
This is because the bit error rate on LTO6 is so high compared to something like the Oracle T10000C tapes. Statistics say you would need two copies of everything on LTO6 to make it as likely to not lose data as a single copy on T10000. We had to do the analysis on the best way to back up 12PB of data, and T10000 won if you were starting from scratch. Since the client already had LTO6 drives (and the silo), that tipped the scales back, but they still have to make two copies and send one offsite to meet the reliability requirements.
Google offers its storage space in return for serving you ads when you use their services.
I haven't ever seen an ad when accessing Google Drive.
I don't see how they can ever serve me an ad on my phone since I use a third-party file manager app that uses the Google Drive API to seamlessly integrate GDrive to my phone file system. On the PC I use a web browser, which can be protected by AdBlock if they ever do start serving ads on the Google Drive page.
Imagine if some poor schumck recorded video on his smartphone of that cop in Ferguson shooting that kid. They'd brick the phone immediately, eliminating the video, and only leaving the schumck's word that he had the video.
How is this any different from the police taking the phone and erasing the video? In order to get the information they need to "brick" the phone, they'd need to look at the phone, thus they would have it in their possession and could do anything nefarious that they chose. The only situation where the government shutting down an individual phone remotely makes any sense is if they believe that phone is a required part of a "people are going to die" plot.
The real problem with a mandated remote kill switch is that every script kiddie on the planet would want to break into whatever controls sending out the "kill" order so they could entertain themselves. With current systems where whatever method I choose to have this functionality is not known, and there are a great many options, it's far less likely that somebody will wipe my phone just for "fun".
Slashdot Mirror
The law is typically written to specifically exempt law enforcement and emergency responders.
Most of those exemptions are only truly valid while on an "emergency" response, even though most cops treat them like full-time exemptions.
Because of the emergency requirement, and because the police car has a radio and most have computers, I can't think of any reason to allow cops to use hand-held cell phones for any reason while driving. If they are responding to an emergency, they should be using official communications methods, and if they aren't, they shouldn't get an exemption from laws.
Why? Really why should I wait? I have an app on my phone that reads my text message if I am in my car. I never respond and it is no more distracting than the radio.
Similarly, my car provides this same feature, and also allows me to send "canned" texts (similar to one-button taunts in online games) with a couple of clicks on the car's touch screen. I also can't edit "regular" texts while the car is in motion, but I can edit either the canned texts to meet my needs or edit a normal text while the car is stopped but still hit "send" while the car is in motion.
I already go to get water, hit the head, or get lunch, several times a day
Damn, how many lunch breaks do you need?
He's a hobbit, so he starts with second breakfast, and it just continues from there.
And if you want to put a 1TB RAM drive into a server?
Buy hardware that supports 2-4TB on the motherboard.
Regardless of the physical size of the server, the price for such a motherboard plus RAM is going to be less than the add-in card plus RAM. It might be a lot more for the add-in solution, as you might not get warranty support unless you buy RAM from the same company that sells the card. You might also have to decide how much RAM disk you want at time of purchase, while system RAM can be upgraded (to the limit of the motherboard) and you can easily decide the size of RAM disk via software.
Also, for Linux, add-in cards might not have drivers, while system RAM does.
Or how about someone starts making RAM drives (for a decent price!) again?
It's easier to just buy more system RAM than to have specialized "RAM drive" hardware.
Then, if caching doesn't give you good enough performance, you can just use software to create a RAM disk.
Your absolutely right. A lot of iPhone users only want an iPhone due to the traditionally tiny form factor. It is somewhat surprising to see Apple alienate this crowd.
There is no reason the iPhone 6 needs to be larger than the iPhone 5, even with a larger screen.
It's just Apple sticking to the exact same space-wasting layout of previous iPhones that makes it so. My phone has a 5.2" 1920x1080 screen, giving it a bigger screen and more resolution than the iPhone 6, yet the iPhone 6 is only 0.25" less wide (and the same height).
Apple design fail.
The iPhone 6 is 4.7", and might fit your hand better than the 4.7" phone you had.
The iPhone 5 has a 5.3" diagonal case with a 4" screen. My LG G2 has a 5.2" diagonal screen with a 5.75" diagonal case.
So, if they lay out the iPhone 6 screen so it uses the phone area better, it should be about a 5.2" screen, which is less than the old iPhones. Even the 6 Plus shouldn't be more than 6" diagonal case, if designed correctly.
When my phone rings, just pulling it out of my pocket to check who's calling is actually kind of a pain in the ass--depending on temp and what gloves I'm wearing, sneaking a peek at my wrist is potentially much easier
A bluetooth headset that speaks caller ID is even easier.
Seriously, if you feel you must have your phone with you while actually on the slope, then you don't understand the concept of "time off". If your clients feel they are so important that not getting back to them in an hour or two will cost you their business, then you are even worse off than someone living paycheck to paycheck.
Can you list what dangerous services are turned on by default on a Windows Server install?
"Server" is set to automatic and shouldn't be required at all unless you want to share files to another machine, and is the historic point of access for most exploits.
"Remote Desktop Services" is set to manual but seems to get started by some other service in a default install.
"WebClient" is set to manual but seems to get started by some other service in a default install, and is dangerous because it allows any program to upload/download files via a "trusted" mechanism, even if the program itself isn't really trustworthy.
I'm sure there are a lot more, but all of my machines have had useless startup services disabled, so I'd have to do a completely fresh install to check for sure.
Prime Video really isn't a good comparison to any other service, because it's really just a free extra, since Prime is mostly about free two-day shipping.
I watch Prime videos for exactly that reason...they are free. The $100 annual fee for Prime is well worth it for my household, as we average one package per week from Amazon.
For example, getting sendmail to not start until the clamd server is ready to accept connections isn't easy using systemd, but trivial using a standard init script.
In sendmail.service:
After=clamd.service
This still doesn't fix the problem I had, which was that the clamd socket wasn't accepting connections when sendmail actually started, so sendmail failed.
Also, you can't edit sendmail.service and have the changes stick, as it is located at /usr/lib/systemd/system/sendmail.service and will be overwritten the next time sendmail is updated. I know there is a poorly documented mechanism for overriding the built-in version of a *.service file, but could never figure out how it really works. OTOH, editing /etc/init.d/sendmail works really well and isn't overwritten by a package update, because packagers know files in /etc/init.d are often changed.
A reasonable dependency system would define exactly what "available" means and have the manager (systemd) run code from the definition that is a test for the required condition.
An example would be to connect to a MySQL database and see if you get a response. A better test would be a user-defined query that should return a known value.
Maybe you haven't heard of the X Window System, introduced over 25 years ago. It allows you to run graphical applications remotely, so the system with the application does not need a GUI at all.
Except that the system with the application still needs 273* extra packages installed to allow it to run X client applications.
That's 273 extra code bases that need to be secured to keep the server safe.
* Although not the actual number, it's not far from the truth.
Chip + PIN effectively mitigates the weakness in magnetic strip data by embedding a chip (physical, something you have) and a pin (something you know) into the transaction process, plus many other security enhancements.
Since some of the cards stolen were debit cards, which require something you have (card with magnetic strip) and something you know (PIN), I don't see how chip+PIN is the holy grail you think it is.
Although there may be more negotiation/handshake at PoS with chip+PIN, it still comes down to two-factor auth to make that sale. And, if somebody can install software/hardware that grabbed mag strip + PIN, they likely can do the same for chip+PIN.
Almost everyone I've asked that has expressed hatred of SystemD hasn't actually used it.
I've used it. I hate it.
Ignoring the very real problem that putting so damn much in PID 1 is dangerous for system stability and security, systemd is generally OK for all distribution-supplied packages. But, if you have anything at all that the packagers didn't think of, it's a pain in the ass. For example, getting sendmail to not start until the clamd server is ready to accept connections isn't easy using systemd, but trivial using a standard init script.
Also, despite the fact that dependencies are baked-in to systemd, it's not at all uncommon for a service that depends on an something else (service, NFS mount, etc.) to still start up before the dependency is fully ready, simply because the default systemd is to assume the dependency is fulfilled as soon as whatever "starts" it returns.
Next, there is no easy way to copy existing dependencies to another service (which would be the best way to start creating your own), mostly because the systemd docs and examples simply suck.
Last, the dependency system absolutely screams for a GUI interface to be able to follow and configure it, but when one finally is created (if it hasn't been already), it'll be useless on servers, because nobody with brains installs a GUI on the server.
The US asked an Irish court to issue a warrant to force production of the data. The Irish court refused to issue the warrant.
And the question here is: Why?
I don't know the laws of Ireland enough to know if the reason the judge refused to issue a warrant is part of the public record. I don't even know if the judge has to articulate a specific reason.
And, although it might be interesting to know, it really doesn't matter. All that matter is that a court with jurisdiction to issue a warrant said "no", and the US tried an end-run around that authority.
But in the end, many companies will choose to foot a larger bill if they can budget for it in advance. "We will need $100,000 for Office licenses" often sounds better than "We didn't budget for it, but it turns out we need $30,000 for various unforeseen development expenses."
It also rarely matters what the total cost turns out to be, but rather the per-seat cost.
If you can get a big enough volume discount, $200 or so per seat doesn't look bad considering that it's a purchase (Office doesn't require yearly fees), and you get a lot more support overall (both official through MS and various help web sites).
Some are freaking-out though, or at least their publicity management people are. From what I gather, someone started tweeting they were fake, and apparently the more were released to prove they weren't.
If all the leaked photos are very "private", then I suspect they are fakes.
On the other hand, if there are lots of "boring" pictures, too, it's likely they are real.
Buy a Winegard 8800 and a Boost XT preamp, and that distance should be no problem.
I have an 8800 (and a 4400) and a low noise pre-amp on each. I didn't say OTA TV was unwatchable...it's actually quite good. Some signals are so strong that the antenna pointed 90 degrees from the source still gets a signal well over 24dB C/N. On the "proper" antenna, signals can hit 30dB C/N.
But, DirecTV still has fewer dropouts than even the strongest channels, since multiple reflected signals can wreak havoc, and stronger signals are often affected even more. I can't control airplanes, moving trees, or other intermittent reflectors, and these happen far more often than really thick clouds, so OTA fares worse than satellite.
And indeed, your usage is typical of any Google Drive user, right?
For phone access, yes, since there are no ads even on the Google-supplied app. The same applies to the custom app for Windows, Mac, etc.
For web access, there aren't currently any ads on the Google Drive page, so, yes, for now. And, if Google changes this, there is always AdBlock.
But if it was an American, residing on European soil, there would be extradition procedures to follow. And those would involve having the local (EU) police generate their own warrant and make their own arrest based upon a formal request.
If you had followed this case, you would know that this is exactly what the US tried to do.
The US asked an Irish court to issue a warrant to force production of the data. The Irish court refused to issue the warrant. So, the US issued a subpoena to Microsoft, who rightly told the US that although the data was on a Microsoft computer, the data was owned by a customer of Microsoft, therefore a warrant would be required. The US court then issued a warrant for Microsoft to produce the data. Microsoft refused, noting that the data was in a foreign country, and warrants are only valid when issued by a court that has jurisdiction over the location of the requested object/data/person. No US court has jurisdiction over Irish soil, thus we end up at today's story.
The actual point of Microsoft's appeal is that the US wants to have a court to be able to issue an order that has the all the advantages of both a warrant and a subpoena, while ignoring their limitatations. The problem with this is that subpoenas are allowed to be fairly vague and apply to anything that is "owned" by the target of the subpoena, regardless of where it is located. Warrants, OTOH, can force the target to hand over something they don't own but over which they have control, but can only request very specific items/data, and have to be issued by a court that has jurisdiction over where the item/data is located.
A short description of what the patch actually is intended to do would not kill Microsoft. I shouldn't have to go hunting for that information if I want it.
In addition, if you have set Windows Update to "download but not install", then it is possible that you don't have Internet access at the time you are thinking of applying the already-downloaded patch.
And LTO is far more reliable than a SATA hard disk.
It depends on what you mean by "reliable". Able to withstand impact, yes. Able to keep bits intact when stored as they should be...both LTO6 and disk are about the same.
This is because the bit error rate on LTO6 is so high compared to something like the Oracle T10000C tapes. Statistics say you would need two copies of everything on LTO6 to make it as likely to not lose data as a single copy on T10000. We had to do the analysis on the best way to back up 12PB of data, and T10000 won if you were starting from scratch. Since the client already had LTO6 drives (and the silo), that tipped the scales back, but they still have to make two copies and send one offsite to meet the reliability requirements.
Google offers its storage space in return for serving you ads when you use their services.
I haven't ever seen an ad when accessing Google Drive.
I don't see how they can ever serve me an ad on my phone since I use a third-party file manager app that uses the Google Drive API to seamlessly integrate GDrive to my phone file system. On the PC I use a web browser, which can be protected by AdBlock if they ever do start serving ads on the Google Drive page.
Imagine if some poor schumck recorded video on his smartphone of that cop in Ferguson shooting that kid. They'd brick the phone immediately, eliminating the video, and only leaving the schumck's word that he had the video.
How is this any different from the police taking the phone and erasing the video? In order to get the information they need to "brick" the phone, they'd need to look at the phone, thus they would have it in their possession and could do anything nefarious that they chose. The only situation where the government shutting down an individual phone remotely makes any sense is if they believe that phone is a required part of a "people are going to die" plot.
The real problem with a mandated remote kill switch is that every script kiddie on the planet would want to break into whatever controls sending out the "kill" order so they could entertain themselves. With current systems where whatever method I choose to have this functionality is not known, and there are a great many options, it's far less likely that somebody will wipe my phone just for "fun".