Bug-Fixing before release/update of any software is always a good idea, be it open source or in properitery software. So I am glad that Linus decided to wait to fix the update before pushing it out.
I think the NVMe issue is a showstopper, which is why they're taking time to fix it. If anyone isn't familiar, NVMe is an SSD attached to the PCIe bus - given we've already maxed out SATA3. NVMe bests that with the newest SSDs pushing 2GB/sec+ in reads and 1GB/sec+ in writes (SATA3 was limited to 540MB/sec, which is why all SSDs pretty much tested at that level).
It sounds like it could be a catastrophic bug and the last thing anyone wants is a kernel release that kills user data
Self-defense is not retribution. Third-party defense is always considered valid when a threat is imminent.
All the data we have shows that devices that are vulnerable to Mirai, et. al. will become Mirai bots in a short amount of time, and will begin attacking third-party Internet infrastructure.
If somebody can show the above claim to be false, please do so, showing reason and evidence.
But in many jurisdictions there can be limits to what you can claim as self-defense. For example, shooting a burglar running away will actually land you with manslaughter or attempted manslaughter charges in quite a few places. The response has to be measured and not excessive.
So depending on where you are, a vulnerable IoT device that getws bricked without being a part of a botnet might be seen as an excessive response, especially if you can do a more measured one instead (e.g., disable routing so it cannot get on the internet, or simply disabling it with a warning). Destroying it or bricking it may be seen as excessive. Now, if it was participating in the botnet, then maybe bricking it can be seen as an appropriate response.
They're adding functionality that Apple refuses to do. If you cheat in a Steam game, your device and account gets banned. On iOS, apparently, you just uninstall and reinstall and then you can fraudlently order cars all over again.
Actually Apple had that ability. The removed it in iOS7 because developers were abusing it for... tracking purposes. They were sending the device unique IDs to advertisers and giving advertisers a per-device view into everything - location information (if allowed), system information, etc.
Apple removed the ability to get that information because it was abused - they now present different forms of unique IDs to apps for various purposes. They have an advertising ID, resettable on user's command and a few others. It is no longer possible to track an individual device because users privacy was being compromised.
So it's not likely it's coming back - developers have shown they cannot be trusted with it.
And if Steam can ban an email and user from their network, so can Uber. Of course, I'm presuming you need an Uber account in order to hail a taxi from them, because they need to charge your credit card for the trip, then there are plenty of ways to track that. Unless a freshly installed Uber only needs a credit card, but I'm sure Uber can track those as well.
And if Uber is using iTunes account balances, then they easy way is to just stop doing that.
That was the case in the US from the Great Depression until the 1990s. Then we repealed the law that required banks to be so boring.
Now banks can invest in derivatives and all sorts of interesting and exciting things. When those exciting investment vehicles turn out to be garbage, we get the 2008 recession
(The 2008 recession in the US was primarily caused by bundled mortgages. Banks and bank-like entities would make a mortgage loan to any vaguely human-like entity that could demonstrate they were alive. Since a large portion of these loans were garbage that would obviously go into default, the banks and bank-like entities bundled them together and then sold "shares" of the bundle. "An individual loan in this bundle may go bad, but surely they won't all go bad!!". To further reassure investors, the bundlers took out insurance policies that were based on other bundled securities not failing. So when the housing market boom inevitably busted, all those bundles turned to shit. And since they were insured via other bundles that had turned to shit, the entire banking sector of the US was in trouble.)
Actually, that happened in the 70s. Mortgage bundling started around that time too, but of course, they only bundled AAA class mortgages together. Which worked until the late 90s or so, when all the AAA mortgages were all bundled together. Banks were happy because this made those investments less boring.
Then someone created a formula that told you how you can combine a bunch of less-than-perfect mortgages (subprime mortgages) and weigh them as if they were AAA mortgages, and banks became happy again because AAA mortgage bundles were boring, and now if they could include AA, A, and lower class mortgages but still value them as AAA mortgages, then it's exciting again.
Until people realized that such mortgage-backed securities, under closer scrutiny were crap because they were backed by crap.
Hell, people were signing up for mortgages that didn't deserve them - there was a nice acronym called NINJA - No Income, No Job Application. Of course that mortgage is going to get defaulted on.
How about just get rid of data caps. My300/100 connection is uncapped with Bell. Why can't it be that way everywhere in Canada?
That's what the CRTC is trying to do with this ruling. By having all traffic count towards your cap, consumers will reasonably demand that their caps be increased. And given that caps are relatively cheap, then raising them costs very little additional money to the iSP.
By doing this ruling, they're making sure users of Netflix etc., who may have been zero rated start demanding that their ISPs give them reasonable caps and not stupidly small ones.
So you're claiming that even where the methodology is faulty, if it differently faulty in an individual case then the person under study must be suspicious?
I don't think you really understand the "faulty" part in "faulty."
Was it faulty methodology, or just unconventional and different? As far as I know (I watched the show) it seemed like a reasonable test that is used for other purposes as well.
And yes, suspicion must be cast. Remember dieselgate? Just because VW cars passed under the standard test meant they passed under a different test. In fact, it was the fact that the test results of the different test didn't line up that caused people to wonder what was happening. And it turns out in the end that the results were being gamed - when the car detects it was being tested, it cheated.
Want another one? Melamine in milk. Chinese farmers were watering down the milk. But if you do that, they can tell because the milk protein concentration goes down as well. So they added melamine to the milk, which resulted in the measured milk protein to be back to normal.
It's entirely possible that Subway is innocent. But it's also just as likely they're cheating. They're well known to abuse their "we're a healthier alternative" to offer pretty lousy food. Heck, for a long time, their "brown bread" (or "whole wheat") actually was white bread colored brown (by the same CBC folks, too). They analyzed the ingredients, and enriched WHITE flour was the first on the list. They found additives like caramel, molasses and others were added to color the bread brown. (Yes, they added a few whole grains in there, after the fact). The reason people found out was diabetics were wondering why after eating a "whole wheat bread" sub from Subway, their blood glucose readings spiked dangerously high - turns out their "brown" bread was basically sugared white bread.
True, the CBC investigation did things in an odd way.
However, the results from the other chicken fell into reasonable expected values (85-95% chicken). Thus, when Subway's fell well outside the expected value, something is up.
Now, granted, using the industry standard testing methods returns the right value, but you do wonder if there's something else going on - is someone gaming the system so it tests properly, or what's happening so that everyone else measures properly
Automatic here, I use the parking brake every time I park. It's the way I was taught to park a car, plus I know it works in the unfortunate event it has to be used as an emergency brake. Car is 40 years old btw.
I was taught the same thing. Yes, you put the car in Park. But you also engage the parking brake because the transmission does lock in Park, but it's only a little piece of metal. The parking brake is cheap, a transmission is expensive.
Also, on modern cars, there is no "e-brake" anymore. The parking brake is just that - a parking brake. You cannot use it for emergency stopping. It activates the rear brakes. The "E-brake" is really just the normal brakes, mostly because modern systems with traction controls, anti-lock brakes, etc, means each wheel gets an independent braking hydraulic line.so taking one out doesn't take out the whole system.
It probably translates all the Linux calls into Windows calls straight into Windows' NTFS driver. So, probably not useful for what you're thinking.
Indeed that's what it is.
WSL is effectively "GNU/kWindows" where Linux ELF binaries can run on the Windows kernel using the Linux kernel personality that translates Linux calls into Windows NT Kernel calls and where security, filesystems, etc are handled by the Windows kernel as expected.
There's no linux code actually in the system (other than perhaps headers translating the syscall numbers into actual system calls). Likewise, networking is done via Windows NDIS networking, as well as all the other kernel services. Several times I had to sit down and figure out what was actually happening - I had to add an/etc/hosts entry and i needed to figure out how it worked. (Hint: WSL is a kernel layer, so what happens is glibc will look at/etc/hosts, so I should edit the ubuntu/etc/hosts, not the Windows one. The Windows one is used by the Win32 resolver, while the Ubuntu one is used by glibc, and the tools I was using use glibc).
But Tesla is in CA, where non-compete agreements are largely void.
Key word: "largely".
If you're a low-level worker, they're basically void - non-competes and non-poaching clauses don't apply.
Non-competes and no-poaching clauses are valid for high-ranking executives though, where it's assumed they are generally intelligent enough to have their own lawyers review and revise contracts and generally have the power on the employment relationship. Plus, the compensation is generally structured around those clauses too.
I still don't get it. What else would you run these apps on if not a Mac or iOS device? (To me, they've always been free so...what changed?)
You don't have to purchase a NEW iOS or Mac to get these apps anymore.
That's what's different. Of course, given that Apple has had this thing going on for years now, I'd be surprised if there was someone that wasn't already eligible for them. You'd have to be toting around a really old iPhone (probably around the 3GS era) or a really old Mac (over 10 years old) to not qualify.
Gerrit requires code be approved before it will merge it into the mainline branches. It replaces a centralized Git server.
Deployments pull from the official Gerrit mainline, while developers can push/pull into their own private branches without requiring approval. But to push to mainline requires approval and review.
And there's a full chain of custody - if some bad code gets approved, you can see all the comments and who approved the change.
It's a bit tricky if you need to revise a fix, but it just means alternate, supported forms of the standard Git commands you're used to.
Or why not remove Burger King from their search engine? A milder version would be pushing up a warning page when searching for Burger King or any of their trademarks...
And the MPAA and RIAA would LOVE this because it means Google CAN do it, WILL do it, and are doing it for stupid reasons.
Instead of having to "legally" prove a site is bad, why not have Google remove piracy sites for possibly having links? I mean, you removed Burger King because they embarrassed you, so why not remove these sites because no proper search engine should link to less than legitimate sites? And BK was for all intents, more legitimate.
As much as Google wants to, they can't, lest they get a flood of requests to ban all sorts of things "because you proven you can, and will do it for the silliest of reasons".
Many people around the world buy their games from US based online stores when that game is not readily available from their local game store... These sales will have been recorded as US sales, even if the product might have ultimately been shipped overseas.
Except I'm fairly certain everywhere the Switch is sold, Nintendo makes sure stores are overstocked with Zelda. Every store I go into has copies of Zelda on the shelf - it's like Nintendo intentionally shipped every store with 10% more copies of Zelda than Switches.
Now, some people might have ordered it ahead of time in case Nintendo short-shipped Zelda, but it appears that no, Nintendo actually flooded stores with copies.
Tune in next week when we discover, again, that some 9-volt batteries have six AAAA-sized cells in them!
Turns out there's a non-destructive way to test for this, too. Though technically, the 9V batteries don't use AAAA sized batteries (they're slightly smaller than AAAA). But it's very close.
The other configuration is flat which is they're put in a plastic tube as a layer cake - a "pile" configuration.
If you have a high-resolution multimeter (6 1/2 digits), observing the terminal voltage and giving the 9V battery a squeeze can tell you the construction. If the voltage increases, it's a pile construction. If it decreases or the slope of the discharge (the multimeter has enough resolution that its input impedance can clearly discharge the battery)
It's one thing to have Apple Watch or FitBit misread your heart rate as happens with tattoos. It's quite another to misread blood sugar and let someone go into shock or worse.
Most people with diabetes that I know are concerned with the inaccuracy of the home kits which works with blood. What do you think happens if they have to sign a EULA giving indemnity to Apple? They won't buy it, and honestly neither would I.
Honestly, we need to see if the tech actually goes anywhere. Then we'll have to wait for the first bad incident due to the tech. My guess? It'll be like an Apple watch. A few people will use it, and not the ones who rely on it.
I suspect it will be a "Not a medical device". As in it will NOT replace a glucosometer and testing strips. Instead, it might be able to give you one of three readings - hyperglycemic, normal, or hypoglycemic. I believe dogs are able to help you with this as well, but they cost a lot of money and require a lot of upkeep that others might not be able to do.
If your watch can warn you that you're getting close to either extreme, there's a chance the user might be able to fix it before collapsing into a coma. Or if the user does collapse, then the watch can immediately place a call to 911 and relay that information to first responders.
It won't replace a meter, but it can potentially warn if the blood sugar is at the extremes as well as provide information to first responders who might not have anyone to ask what's going on. The watch itself can be screaming out "Diabetic - hypoglycemic" on the screen which will aid them in reviving you.
I use Paypal and then after I pass that, I deactivate the Paypal authorization. Once deactivated, that link cannot be used to charge your account anymore. The only way to reactivate it is create a new link which requires logging into your paypal account again.
The standalone version works quite well - plug in a controller, plug in the TV, plug in power, and game away. It's super simple to use.
Add in all the network connectivity and expandability and you've complicated it 100 times. Plus it has to cost more - additional storage, network/wifi equipment, etc. And then there's the whole billing and account stuff that needs to be done.
The NES Classic as it is right now is perfect. You can be playing games in about 30 seconds. An enhanced version will cost more and be less easy to use.
I doubt it. Because I would bet if they changed classes to 11AM then students would just stay up for 3 hours later knowing they could sleep in. Once their bodies adjusted it would business as usual. Waking up groggy for 11AM classes instead.
You'd think that, but no. For teenagers, their circadian rhythm is offset a few hours so a later start time improves matters, even when you realize they sleep later. There is a lot of truth to the teenager sleeping in beyond noon on a weekend.
The big reason we don't start high school around 10AM or so is because it would end late for teens who might have a job, sports, etc.
and I seriously doubt anyone will ever update the system to 64 bits.
One of the reasons why OS/2 runs so well on x86 is it's heavily tied to the i386 CPU. It's one of the few OSes that actually uses multiple rings of protection. Linux, NT, etc, use ring0 and ring3 (for "kernel" and "user" mode) while OS/2 uses ring0 for kernel and device drivers, ring2 for privileged applications, and ring3 for everything else.
And it's also very adept - driver support for OS/2 comes in both 16 and 32 bit varieties - the 16 bit drivers ran alongside 32-bit drivers with some tricky work, and even able to run DOS drivers in a VDM without stalling out like Windows did (Windows using a DOS driver would grab a giant kernel lock, go single tasking and then run the DOS code. It's why formatting a floppy disk on Windows 9x was frustrating as Windows went to DOS for that. OS/2 didn't do any of that and used a pre-emptively multitasked DOS box instead.
It's why it had excellent driver support through the years because no driver ever got deprecated - OS/2 simply ran the driver in an appropriate environment
M$ is making tons of cash from Android device manufacturers, with the help of a secret set of patents. May be Google is trying to bring together a set as big as M$ one ?
Any whistleblower around to finally show us what M$ patents are ?
exFAT was one of them, vFAT was another.
But the reason Microsoft doesn't want to go after Google is they can't - Google doesn't use any of that technology. The FAT patents aren't needed if you're not using FAT at all - hence the lack of an SD card support in the Nexus and Pixel phones. And you can eliminate any need for FAT if you use MTP as your USB transfer protocol that Windows supports.
I'm sure more reasons are simple - all the other phone vendors are trying hard to make their phones not appear all the same at the store. You used to be able to do this by using custom launchers and shells and skins, but Google outlawed them (HTC Sense, TouchWiz, etc) so now they have to add features. And some vendors have interesting features - dual cameras, etc. If Google wants to integrate support into the mainline, they'll need patent licenses to do so, hence this agreement.
You have to remember TLS was (along with Tron) the pioneers of CGI - and that computers in the 80s were kinda weak. That said, they did render TLS on a Cray X-MP, the worlds fastest computer back in 1983 running at a blistering 115MHz with around 16MB of RAM. (Though you could get large storage arrays, even 1GB SSDs). Even then at 250k polygons it still was a challenge for the machine. Plus the technology had to be invented - unlike today where you can go out and download yourself a high quality renderer (RenderMan free) as well as get access to high quality tools, back then all had to be created, on production timelines.
Plus, I don't believe texturing was available back then - everything CGI was simply shaded...
When I advertise for a "programmer", I usually don't have a specific salary in mind. If one applicant is more capable than another, then I will offer more. If I put a low salary range in the ad, the better candidates will not apply. If I put a high range, then I will be flooded with responses from lousy candidates that are not even remotely qualified for that salary.
The problem is the employer-employee relationship is asymmetrical. The employer has all the power.
In your case, you should have a range of salary in mind for the position - either based on industry averages, or what you think someone with the skill set you need should be paid. To argue otherwise is disingenuous.
If you advertise low, yes, the better candidates won't apply - they assume you want someone much more junior. If you advertise high, then yes you get more crap, but you can filter them out - if you need N years of experience and the candidate doesn't have it, then either you need wiggle room on the requirement (and thus lower than the range given) or you drop the candidate for not being able to read and meet the requirements.
Or you can waste everyone's time and interview them then make offers.
Right now, asking interviewees their current salary is like telling employees to keep their raises and current salaries private - it only benefits the employer when people are left guessing. (Many people comply because they worry they're overpaid. The reality so far is everywhere it's happened, the underpaid are the ones who benefit the most.
It should be noted that the "City of London" is London itself and older than antiquity. What we know of London today was where the political powers were removed from the City of London to Westminster and thus "London" grew around it.
Incompatible = Apple deciding to block fingerprint readers with a different ID than originally came with the phone. A security move that only makes sense during the initial design - not when done after the phone is out there. It was a valid repair and the iPhone offers no way to pair with a new fingerprint reader except by Apple (which is just as bad as putting a chip on a printer cartridge and should be illegal).
Or how about to easily prevent hacking via the fingerprint sensor?
Consider this scenario - FBI acquires your phone. They remove the display and replace the touch sensor with their own special fingerprint sensor they designed. They they go "hey Apple, I just replaced the screen on my phone, care to re-pair the fingerprint sensor?", thus putting their fake fingerprint sensor on your phone. Their fake sensor then proceeds to hack until it can unlock your phone.
Without allowing easy re-pairing, the fake sensor data is simply ignored - the secure enclave knows someone tampered with the sensor and can consider all data being sent to it as attempts to hack in. Of course, the proper way to do it is to re-pair the sensor and secure enclave only after a full system wipe and restore.
Of course, it's also a given Apple didn't know how many people got affected by this - all the people who brought in their phones had the repair done properly and there was no sense of the scale to which people were using third party unauthorized repair shops.who were not doing the proper steps.
Slashdot Mirror
I think the NVMe issue is a showstopper, which is why they're taking time to fix it. If anyone isn't familiar, NVMe is an SSD attached to the PCIe bus - given we've already maxed out SATA3. NVMe bests that with the newest SSDs pushing 2GB/sec+ in reads and 1GB/sec+ in writes (SATA3 was limited to 540MB/sec, which is why all SSDs pretty much tested at that level).
It sounds like it could be a catastrophic bug and the last thing anyone wants is a kernel release that kills user data
But in many jurisdictions there can be limits to what you can claim as self-defense. For example, shooting a burglar running away will actually land you with manslaughter or attempted manslaughter charges in quite a few places. The response has to be measured and not excessive.
So depending on where you are, a vulnerable IoT device that getws bricked without being a part of a botnet might be seen as an excessive response, especially if you can do a more measured one instead (e.g., disable routing so it cannot get on the internet, or simply disabling it with a warning). Destroying it or bricking it may be seen as excessive. Now, if it was participating in the botnet, then maybe bricking it can be seen as an appropriate response.
Actually Apple had that ability. The removed it in iOS7 because developers were abusing it for... tracking purposes. They were sending the device unique IDs to advertisers and giving advertisers a per-device view into everything - location information (if allowed), system information, etc.
Apple removed the ability to get that information because it was abused - they now present different forms of unique IDs to apps for various purposes. They have an advertising ID, resettable on user's command and a few others. It is no longer possible to track an individual device because users privacy was being compromised.
So it's not likely it's coming back - developers have shown they cannot be trusted with it.
And if Steam can ban an email and user from their network, so can Uber. Of course, I'm presuming you need an Uber account in order to hail a taxi from them, because they need to charge your credit card for the trip, then there are plenty of ways to track that. Unless a freshly installed Uber only needs a credit card, but I'm sure Uber can track those as well.
And if Uber is using iTunes account balances, then they easy way is to just stop doing that.
Actually, that happened in the 70s. Mortgage bundling started around that time too, but of course, they only bundled AAA class mortgages together. Which worked until the late 90s or so, when all the AAA mortgages were all bundled together. Banks were happy because this made those investments less boring.
Then someone created a formula that told you how you can combine a bunch of less-than-perfect mortgages (subprime mortgages) and weigh them as if they were AAA mortgages, and banks became happy again because AAA mortgage bundles were boring, and now if they could include AA, A, and lower class mortgages but still value them as AAA mortgages, then it's exciting again.
Until people realized that such mortgage-backed securities, under closer scrutiny were crap because they were backed by crap.
Hell, people were signing up for mortgages that didn't deserve them - there was a nice acronym called NINJA - No Income, No Job Application. Of course that mortgage is going to get defaulted on.
That's what the CRTC is trying to do with this ruling. By having all traffic count towards your cap, consumers will reasonably demand that their caps be increased. And given that caps are relatively cheap, then raising them costs very little additional money to the iSP.
By doing this ruling, they're making sure users of Netflix etc., who may have been zero rated start demanding that their ISPs give them reasonable caps and not stupidly small ones.
Was it faulty methodology, or just unconventional and different? As far as I know (I watched the show) it seemed like a reasonable test that is used for other purposes as well.
And yes, suspicion must be cast. Remember dieselgate? Just because VW cars passed under the standard test meant they passed under a different test. In fact, it was the fact that the test results of the different test didn't line up that caused people to wonder what was happening. And it turns out in the end that the results were being gamed - when the car detects it was being tested, it cheated.
Want another one? Melamine in milk. Chinese farmers were watering down the milk. But if you do that, they can tell because the milk protein concentration goes down as well. So they added melamine to the milk, which resulted in the measured milk protein to be back to normal.
It's entirely possible that Subway is innocent. But it's also just as likely they're cheating. They're well known to abuse their "we're a healthier alternative" to offer pretty lousy food. Heck, for a long time, their "brown bread" (or "whole wheat") actually was white bread colored brown (by the same CBC folks, too). They analyzed the ingredients, and enriched WHITE flour was the first on the list. They found additives like caramel, molasses and others were added to color the bread brown. (Yes, they added a few whole grains in there, after the fact). The reason people found out was diabetics were wondering why after eating a "whole wheat bread" sub from Subway, their blood glucose readings spiked dangerously high - turns out their "brown" bread was basically sugared white bread.
True, the CBC investigation did things in an odd way.
However, the results from the other chicken fell into reasonable expected values (85-95% chicken). Thus, when Subway's fell well outside the expected value, something is up.
Now, granted, using the industry standard testing methods returns the right value, but you do wonder if there's something else going on - is someone gaming the system so it tests properly, or what's happening so that everyone else measures properly
I was taught the same thing. Yes, you put the car in Park. But you also engage the parking brake because the transmission does lock in Park, but it's only a little piece of metal. The parking brake is cheap, a transmission is expensive.
Also, on modern cars, there is no "e-brake" anymore. The parking brake is just that - a parking brake. You cannot use it for emergency stopping. It activates the rear brakes. The "E-brake" is really just the normal brakes, mostly because modern systems with traction controls, anti-lock brakes, etc, means each wheel gets an independent braking hydraulic line.so taking one out doesn't take out the whole system.
Indeed that's what it is.
WSL is effectively "GNU/kWindows" where Linux ELF binaries can run on the Windows kernel using the Linux kernel personality that translates Linux calls into Windows NT Kernel calls and where security, filesystems, etc are handled by the Windows kernel as expected.
There's no linux code actually in the system (other than perhaps headers translating the syscall numbers into actual system calls). Likewise, networking is done via Windows NDIS networking, as well as all the other kernel services. Several times I had to sit down and figure out what was actually happening - I had to add an /etc/hosts entry and i needed to figure out how it worked. (Hint: WSL is a kernel layer, so what happens is glibc will look at /etc/hosts, so I should edit the ubuntu /etc/hosts, not the Windows one. The Windows one is used by the Win32 resolver, while the Ubuntu one is used by glibc, and the tools I was using use glibc).
Key word: "largely".
If you're a low-level worker, they're basically void - non-competes and non-poaching clauses don't apply.
Non-competes and no-poaching clauses are valid for high-ranking executives though, where it's assumed they are generally intelligent enough to have their own lawyers review and revise contracts and generally have the power on the employment relationship. Plus, the compensation is generally structured around those clauses too.
You don't have to purchase a NEW iOS or Mac to get these apps anymore.
That's what's different. Of course, given that Apple has had this thing going on for years now, I'd be surprised if there was someone that wasn't already eligible for them. You'd have to be toting around a really old iPhone (probably around the 3GS era) or a really old Mac (over 10 years old) to not qualify.
Exactly.
Gerrit requires code be approved before it will merge it into the mainline branches. It replaces a centralized Git server.
Deployments pull from the official Gerrit mainline, while developers can push/pull into their own private branches without requiring approval. But to push to mainline requires approval and review.
And there's a full chain of custody - if some bad code gets approved, you can see all the comments and who approved the change.
It's a bit tricky if you need to revise a fix, but it just means alternate, supported forms of the standard Git commands you're used to.
And the MPAA and RIAA would LOVE this because it means Google CAN do it, WILL do it, and are doing it for stupid reasons.
Instead of having to "legally" prove a site is bad, why not have Google remove piracy sites for possibly having links? I mean, you removed Burger King because they embarrassed you, so why not remove these sites because no proper search engine should link to less than legitimate sites? And BK was for all intents, more legitimate.
As much as Google wants to, they can't, lest they get a flood of requests to ban all sorts of things "because you proven you can, and will do it for the silliest of reasons".
Except I'm fairly certain everywhere the Switch is sold, Nintendo makes sure stores are overstocked with Zelda. Every store I go into has copies of Zelda on the shelf - it's like Nintendo intentionally shipped every store with 10% more copies of Zelda than Switches.
Now, some people might have ordered it ahead of time in case Nintendo short-shipped Zelda, but it appears that no, Nintendo actually flooded stores with copies.
Turns out there's a non-destructive way to test for this, too. Though technically, the 9V batteries don't use AAAA sized batteries (they're slightly smaller than AAAA). But it's very close.
The other configuration is flat which is they're put in a plastic tube as a layer cake - a "pile" configuration.
If you have a high-resolution multimeter (6 1/2 digits), observing the terminal voltage and giving the 9V battery a squeeze can tell you the construction. If the voltage increases, it's a pile construction. If it decreases or the slope of the discharge (the multimeter has enough resolution that its input impedance can clearly discharge the battery)
EEV Blog #518 - https://www.eevblog.com/2013/0...
EEV B;pcl #515 - https://www.youtube.com/watch?...
I suspect it will be a "Not a medical device". As in it will NOT replace a glucosometer and testing strips. Instead, it might be able to give you one of three readings - hyperglycemic, normal, or hypoglycemic. I believe dogs are able to help you with this as well, but they cost a lot of money and require a lot of upkeep that others might not be able to do.
If your watch can warn you that you're getting close to either extreme, there's a chance the user might be able to fix it before collapsing into a coma. Or if the user does collapse, then the watch can immediately place a call to 911 and relay that information to first responders.
It won't replace a meter, but it can potentially warn if the blood sugar is at the extremes as well as provide information to first responders who might not have anyone to ask what's going on. The watch itself can be screaming out "Diabetic - hypoglycemic" on the screen which will aid them in reviving you.
I use Paypal and then after I pass that, I deactivate the Paypal authorization. Once deactivated, that link cannot be used to charge your account anymore. The only way to reactivate it is create a new link which requires logging into your paypal account again.
I don't know.
The standalone version works quite well - plug in a controller, plug in the TV, plug in power, and game away. It's super simple to use.
Add in all the network connectivity and expandability and you've complicated it 100 times. Plus it has to cost more - additional storage, network/wifi equipment, etc. And then there's the whole billing and account stuff that needs to be done.
The NES Classic as it is right now is perfect. You can be playing games in about 30 seconds. An enhanced version will cost more and be less easy to use.
You'd think that, but no. For teenagers, their circadian rhythm is offset a few hours so a later start time improves matters, even when you realize they sleep later. There is a lot of truth to the teenager sleeping in beyond noon on a weekend.
The big reason we don't start high school around 10AM or so is because it would end late for teens who might have a job, sports, etc.
One of the reasons why OS/2 runs so well on x86 is it's heavily tied to the i386 CPU. It's one of the few OSes that actually uses multiple rings of protection. Linux, NT, etc, use ring0 and ring3 (for "kernel" and "user" mode) while OS/2 uses ring0 for kernel and device drivers, ring2 for privileged applications, and ring3 for everything else.
And it's also very adept - driver support for OS/2 comes in both 16 and 32 bit varieties - the 16 bit drivers ran alongside 32-bit drivers with some tricky work, and even able to run DOS drivers in a VDM without stalling out like Windows did (Windows using a DOS driver would grab a giant kernel lock, go single tasking and then run the DOS code. It's why formatting a floppy disk on Windows 9x was frustrating as Windows went to DOS for that. OS/2 didn't do any of that and used a pre-emptively multitasked DOS box instead.
It's why it had excellent driver support through the years because no driver ever got deprecated - OS/2 simply ran the driver in an appropriate environment
exFAT was one of them, vFAT was another.
But the reason Microsoft doesn't want to go after Google is they can't - Google doesn't use any of that technology. The FAT patents aren't needed if you're not using FAT at all - hence the lack of an SD card support in the Nexus and Pixel phones. And you can eliminate any need for FAT if you use MTP as your USB transfer protocol that Windows supports.
I'm sure more reasons are simple - all the other phone vendors are trying hard to make their phones not appear all the same at the store. You used to be able to do this by using custom launchers and shells and skins, but Google outlawed them (HTC Sense, TouchWiz, etc) so now they have to add features. And some vendors have interesting features - dual cameras, etc. If Google wants to integrate support into the mainline, they'll need patent licenses to do so, hence this agreement.
You have to remember TLS was (along with Tron) the pioneers of CGI - and that computers in the 80s were kinda weak. That said, they did render TLS on a Cray X-MP, the worlds fastest computer back in 1983 running at a blistering 115MHz with around 16MB of RAM. (Though you could get large storage arrays, even 1GB SSDs). Even then at 250k polygons it still was a challenge for the machine. Plus the technology had to be invented - unlike today where you can go out and download yourself a high quality renderer (RenderMan free) as well as get access to high quality tools, back then all had to be created, on production timelines.
Plus, I don't believe texturing was available back then - everything CGI was simply shaded...
The problem is the employer-employee relationship is asymmetrical. The employer has all the power.
In your case, you should have a range of salary in mind for the position - either based on industry averages, or what you think someone with the skill set you need should be paid. To argue otherwise is disingenuous.
If you advertise low, yes, the better candidates won't apply - they assume you want someone much more junior. If you advertise high, then yes you get more crap, but you can filter them out - if you need N years of experience and the candidate doesn't have it, then either you need wiggle room on the requirement (and thus lower than the range given) or you drop the candidate for not being able to read and meet the requirements.
Or you can waste everyone's time and interview them then make offers.
Right now, asking interviewees their current salary is like telling employees to keep their raises and current salaries private - it only benefits the employer when people are left guessing. (Many people comply because they worry they're overpaid. The reality so far is everywhere it's happened, the underpaid are the ones who benefit the most.
CGP Grey has more information about the City of London
https://www.youtube.com/playli...
It should be noted that the "City of London" is London itself and older than antiquity. What we know of London today was where the political powers were removed from the City of London to Westminster and thus "London" grew around it.
Or how about to easily prevent hacking via the fingerprint sensor?
Consider this scenario - FBI acquires your phone. They remove the display and replace the touch sensor with their own special fingerprint sensor they designed. They they go "hey Apple, I just replaced the screen on my phone, care to re-pair the fingerprint sensor?", thus putting their fake fingerprint sensor on your phone. Their fake sensor then proceeds to hack until it can unlock your phone.
Without allowing easy re-pairing, the fake sensor data is simply ignored - the secure enclave knows someone tampered with the sensor and can consider all data being sent to it as attempts to hack in. Of course, the proper way to do it is to re-pair the sensor and secure enclave only after a full system wipe and restore.
Of course, it's also a given Apple didn't know how many people got affected by this - all the people who brought in their phones had the repair done properly and there was no sense of the scale to which people were using third party unauthorized repair shops.who were not doing the proper steps.