~ $ df -h
Filesystem Size Used Avail Use% Mounted on /dev/sda1 292G 53G 225G 19%/
Hm...:|
Bah.
# vgdisplay 5g
[...]
Free PE / Size 419390 / 1.60 TiB
# lvcreate -n wiki -L 1T 5g
Logical volume "wiki" created
# mkfs -t xfs/dev/5g/wiki
[...]
# mount/dev/5g/wiki/mnt
I started the torrents, but so far I'm only getting about 3 MBps. My connection should be able to manage close to 10 MBps, so hopefully that will pick up.
That's one explanation, but it doesn't hold up well at Google, because all engineering managers are engineers, all the way up to (and including) the CEO. Here's another, simpler explanation: The article is wrong.
I also work for Google, and the parent reflects what I see as well.
Relatively few engineers have a 20% project, but it's really their choice. Management is supportive of 20% time. FWIW, of the engineers who sit near me, 40% have 20% projects (including me). On my team, 100% have 20% projects -- but there are only two of us.
And since I'm a few years older than Vince Vaughan, I seriously doubt I'd quite fit in anymore.
Google isn't so young any more. I'm 44 and I fit right in. Some of the engineers I work with are in their 50s and 60s. There are also plenty of younger engineers, of course, but I'd say the median age is around 35, at least in my neck of the woods.
if you were in a position to look at the requests and their rationale, you'd agree
If you were a dog, and thought like a dog, you would behave like a dog.
Allow me to rephrase: I think just about any intelligent, reasonable person would look at the warrants in criminal investigations, the subpoenas in civil suits, etc., and find the requests reasonable, appropriate and in the interest of justice and society in general. You know, the 4th amendment allows warrants for a reason... because they make society a better place. National Security Letters... that I'm not so sure about. We need real oversight, and (as mentioned in another article on/. today) we don't have it. But, given appropriate oversight to make sure they're really justified -- meaning that they're directed very specifically at people for whom there is convincing evidence of terrorist or other activity that endangers large numbers of lives -- even NSLs are probably a good thing.
Short story: the sooner we expand our lives, the better, as we can sustain doubling the population _now_, but that might not be the case after we travel further along the growth curve.
Actually, based on current trends, we're not going to travel much further along the growth curve.
The global birth rate has peaked and flattened, and I'm talking about rate in number of children per year, not number of children per woman. The population is still growing because that peak rate is much higher than it was. But at this point, we're basically getting two billion new people in each generation, and we live about five generations, so we're on course to max out at 10 billion people. The developed world is already at or below replacement rate (including the US, but immigration offsets the losses and keeps us growing) and the developing world's birth rates are declining as they get wealthier. So in all probability, after we hit that peak population, the total will begin to fall.
Now, if you throw a doubled lifespan in there, then we are going to end up at a much higher number. Probably not 20 billion, but higher. Assuming you could pick a time, it would be better to wait until the population peaks and then falls some before increasing life span.
Once Google unequivocally tells the feds to fuck off the next time they come sniffing around for user data, I'll put some stock into such supposed privacy measures.
Google refuses ~30% of government requests for user data.
Keep in mind that most requests are subpoenas (which can only get extremely limited data; name and IP address, basically), court orders (which can get a bit more, but not e-mail contents) and search warrants, and I think it's quite likely that if you were in a position to look at the requests and their rationale, you'd agree that most of them are legitimate and not only legally must be respected, but should be respected, because it's the right thing to do.
there is no such thing as profit on hardware sales
Apple disagrees.
Re:Thanks, from an embedded designer.
on
Debian Turns 20
·
· Score: 1
Some of the devices he programs for only has 4-16MB of RAM for everything.
LOL.
My major embedded experience was on some control boards with 4 MiB RAM and 4 MiB Flash ROM. That was pretty beefy, though, and had a full-featured, pre-emptive multi-tasking OS (VxWorks). The really interesting work was on the five V25 microcontrollers which we used to run the many serial ports (actually, there were UARTs to actually run the serial ports, but the V25s were responsible for the care and feeding of the UARTs). Each V25 had 16 KiB of RAM, no ROM. There was no OS at all on them; it was bare metal programming.
Re:most memorable and significant fork
on
Debian Turns 20
·
· Score: 1
But you mentioning Knoppix... Didn't they get the whole running a distro from a CD only into the big time?
Yep, LiveCD operation was Knoppix' raison d'etre. It was an awesome system repair tool back in the day. System borked? Boot from your trusty Knoppix CD (later, USB stick), mount the drive and fix it. A few years later I switched to Damn Small Linux for that purpose instead, but still kept a Knoppix CD around.
I thought 1.0 versions of anything were never released, firstly because they usually suck and secondly because it looks better form a marketing perspective to start around 3 or 4.
In the open source world, historically the problem has been the opposite. Many packages have version 0.1, 0.5, 0.6, 0.8, 0.9, 0.91, 0.92... and so on. There's a reluctance to apply the "1.0" label because that means you have something that's really "done" in some sense.
Turn-by-turn directions that appear to be floating in the air 8 feet in front of you, a little up and to the right so they're out of your central field of vision, seem like a safer option than putting the same directions on a screen in the center of the console and much safer than on a little handheld screen.
Short of an actual HUD, Glass seems like the ideal way to display driving-related information. In theory, at least. I've read that the current generation isn't quite bright enough, so directions are hard to see. That may be fixed in the public release model, dunno.
Of course, people can (and some will) use Glass for other, distracting, purposes while driving. But that's hardly the fault of the technology.
Google has said it's not going to allow advertising on Glass. I think the business model for Glass is just profit on hardware sales. I don't actually know the plan, that's just a guess.
Oh, and regarding the other likely assumption of evil that I know someone is going to post: Google is also not going to be streaming everything your Glass sees to their servers. Privacy issues aside, it'd destroy the battery life and blow through your mobile data plan in no time.
(Disclaimer: I work for Google but have no inside knowledge of Glass.)
True random numbers are as simple as a reversed Zener diode connected to an A/D converter... quantum tunneling across the diode creates truly random signal, equivalent to thermal noise.
Sure, if you don't mind highly-exploitable biases in your random numbers. Randomness is not enough. You also need to obtain a uniform distribution. Simple hardware designs may also be influenced by external signals, which may be predictable or even controllable by the attacker. The fact that there is some quantum-derived randomness doesn't mean there is no predictability.
Doing this stuff right is hard./dev/random in the Linux kernel actually does a fine job, better than any cost-effective hardware implementation will. Though, a cheap and biased hardware RNG can be used to help feed the kernel's entropy pool.
IMO it also depends heavily on the nature of the course.
I really suck at memorizing anything, whether I write it down or not. Though writing it helps, assuming the process of writing doesn't cause me to miss something else I should have learned. That's basically why I ended up with a math degree, because in math courses (especially upper division) there is basically no point in trying to memorize anything. Either you understand it or you've got nothing, and what little you might have to memorize is all in the book anyway, in bolded and bulleted definitions and theorems. In those classes it's clearly far better to focus on listening, thinking and understanding; taking notes is a waste of time and a pure distraction.
In classes which are all about absorbing a big pile of facts, memorization aids are useful. In classes which are about understanding ideas, memorization is useless and so are notes. Courses which mix facts and ideas need a mixed approach, and are often the most difficult.
The best example I've seen of the latter is a freshman-level physics (mechanics, mostly) class my wife took, which didn't assume, nor explicate, any calculus. That class was a cast-iron bitch, because if you don't understand simple calculus you have to memorize a whole bunch of variations of every fundamental equation. For her final she had literally pages of equations to memorize, with no obvious relationship between expressions which are actually deeply related, plus she had to understand what all of them meant. The result was a class that was an order of magnitude more difficult than the same class with a calculus pre-req. I tried to get her to let me teach her the necessary calculus, promising that it wasn't more stuff to learn, but less. She didn't believe me and just gutted the course out as presented.
In a few years, perhaps technology will allow us to flip the problem. Rather than entering everything into the EMR and printing it out to have paper, you might be able to use paper as the data entry tool. If you could easily scan the paper and automatically extract the details needed to update the EMR, then you could have both. Of course, accurate automated analysis of doctors' handwriting is just a wee bit challenging. But machines should ultimately be able to do it as well as people, including being able to ask when they just can't make it out.
When those workers get to be about 35 YO, they'll be back to reality when they're looking for work and a place to live that they can afford.
Enjoy it while you can - your ass will be kicked to the curb before you know it.
Not at Google. Google employs lots of older engineers. I was hired at age 41 and I'm far from the oldest around. One of the guys I work with is in his 60s. The company still tends to be skewed towards the younger end of the demographics, but I'll bet the median age is in the low to mid-30s and rising. The Colorado office, where I work, tends to be older still, because the cost of living in the bay area drives away people with families.
Bear in mind that most Google employees are not "techies". They're sales reps selling ads. When you think Google, think "Mad Men", not rocket science.
Um, roughly 50% of Google's 40,000 full-time employees are software engineers. The other half make up everything else. I don't know numbers for sales reps, but I'd be surprised if they comprised 20% of the FTEs.
Oh, it's also worth pointing out that Wickard v. Fillburn was a direct result of FDR's court packing proposal. Like most of the related rulings of the time, it's bad law. Not that I see any realistic chance of reversing it any time soon.
Yes, I understand that's the position the courts have taken. The result is to destroy the notion that the constitution limits the scope of the federal government in any way, since absolutely anything can potentially affect commerce.
This is utter nonsense - the US government is allowed to regulate anyone conducting financial transactions within the US.
You mean interstate financial transactions, of course.
However, this article isn't about the US government, it's about New York state, which really does have legal authority to regulate any financial transactions in New York.
Slashdot Mirror
~ $ df -h
/dev/sda1 292G 53G 225G 19% /
Filesystem Size Used Avail Use% Mounted on
Hm... :|
Bah.
# vgdisplay 5g /dev/5g/wiki /dev/5g/wiki /mnt
[...]
Free PE / Size 419390 / 1.60 TiB
# lvcreate -n wiki -L 1T 5g
Logical volume "wiki" created
# mkfs -t xfs
[...]
# mount
I started the torrents, but so far I'm only getting about 3 MBps. My connection should be able to manage close to 10 MBps, so hopefully that will pick up.
Yes, this is my home file server :)
That's one explanation, but it doesn't hold up well at Google, because all engineering managers are engineers, all the way up to (and including) the CEO. Here's another, simpler explanation: The article is wrong.
I also work for Google, and the parent reflects what I see as well.
Relatively few engineers have a 20% project, but it's really their choice. Management is supportive of 20% time. FWIW, of the engineers who sit near me, 40% have 20% projects (including me). On my team, 100% have 20% projects -- but there are only two of us.
And since I'm a few years older than Vince Vaughan, I seriously doubt I'd quite fit in anymore.
Google isn't so young any more. I'm 44 and I fit right in. Some of the engineers I work with are in their 50s and 60s. There are also plenty of younger engineers, of course, but I'd say the median age is around 35, at least in my neck of the woods.
if you were in a position to look at the requests and their rationale, you'd agree
If you were a dog, and thought like a dog, you would behave like a dog.
Allow me to rephrase: I think just about any intelligent, reasonable person would look at the warrants in criminal investigations, the subpoenas in civil suits, etc., and find the requests reasonable, appropriate and in the interest of justice and society in general. You know, the 4th amendment allows warrants for a reason... because they make society a better place. National Security Letters... that I'm not so sure about. We need real oversight, and (as mentioned in another article on /. today) we don't have it. But, given appropriate oversight to make sure they're really justified -- meaning that they're directed very specifically at people for whom there is convincing evidence of terrorist or other activity that endangers large numbers of lives -- even NSLs are probably a good thing.
Short story: the sooner we expand our lives, the better, as we can sustain doubling the population _now_, but that might not be the case after we travel further along the growth curve.
Actually, based on current trends, we're not going to travel much further along the growth curve.
The global birth rate has peaked and flattened, and I'm talking about rate in number of children per year, not number of children per woman. The population is still growing because that peak rate is much higher than it was. But at this point, we're basically getting two billion new people in each generation, and we live about five generations, so we're on course to max out at 10 billion people. The developed world is already at or below replacement rate (including the US, but immigration offsets the losses and keeps us growing) and the developing world's birth rates are declining as they get wealthier. So in all probability, after we hit that peak population, the total will begin to fall.
Now, if you throw a doubled lifespan in there, then we are going to end up at a much higher number. Probably not 20 billion, but higher. Assuming you could pick a time, it would be better to wait until the population peaks and then falls some before increasing life span.
Once Google unequivocally tells the feds to fuck off the next time they come sniffing around for user data, I'll put some stock into such supposed privacy measures.
Google refuses ~30% of government requests for user data.
Keep in mind that most requests are subpoenas (which can only get extremely limited data; name and IP address, basically), court orders (which can get a bit more, but not e-mail contents) and search warrants, and I think it's quite likely that if you were in a position to look at the requests and their rationale, you'd agree that most of them are legitimate and not only legally must be respected, but should be respected, because it's the right thing to do.
http://www.google.com/transparencyreport/userdatarequests/
http://www.google.com/transparencyreport/userdatarequests/legalprocess/
there is no such thing as profit on hardware sales
Apple disagrees.
Some of the devices he programs for only has 4-16MB of RAM for everything.
LOL.
My major embedded experience was on some control boards with 4 MiB RAM and 4 MiB Flash ROM. That was pretty beefy, though, and had a full-featured, pre-emptive multi-tasking OS (VxWorks). The really interesting work was on the five V25 microcontrollers which we used to run the many serial ports (actually, there were UARTs to actually run the serial ports, but the V25s were responsible for the care and feeding of the UARTs). Each V25 had 16 KiB of RAM, no ROM. There was no OS at all on them; it was bare metal programming.
But you mentioning Knoppix... Didn't they get the whole running a distro from a CD only into the big time?
Yep, LiveCD operation was Knoppix' raison d'etre. It was an awesome system repair tool back in the day. System borked? Boot from your trusty Knoppix CD (later, USB stick), mount the drive and fix it. A few years later I switched to Damn Small Linux for that purpose instead, but still kept a Knoppix CD around.
I thought 1.0 versions of anything were never released, firstly because they usually suck and secondly because it looks better form a marketing perspective to start around 3 or 4.
In the open source world, historically the problem has been the opposite. Many packages have version 0.1, 0.5, 0.6, 0.8, 0.9, 0.91, 0.92... and so on. There's a reluctance to apply the "1.0" label because that means you have something that's really "done" in some sense.
Turn-by-turn directions that appear to be floating in the air 8 feet in front of you, a little up and to the right so they're out of your central field of vision, seem like a safer option than putting the same directions on a screen in the center of the console and much safer than on a little handheld screen.
Short of an actual HUD, Glass seems like the ideal way to display driving-related information. In theory, at least. I've read that the current generation isn't quite bright enough, so directions are hard to see. That may be fixed in the public release model, dunno.
Of course, people can (and some will) use Glass for other, distracting, purposes while driving. But that's hardly the fault of the technology.
Google has said it's not going to allow advertising on Glass. I think the business model for Glass is just profit on hardware sales. I don't actually know the plan, that's just a guess.
Oh, and regarding the other likely assumption of evil that I know someone is going to post: Google is also not going to be streaming everything your Glass sees to their servers. Privacy issues aside, it'd destroy the battery life and blow through your mobile data plan in no time.
(Disclaimer: I work for Google but have no inside knowledge of Glass.)
True random numbers are as simple as a reversed Zener diode connected to an A/D converter... quantum tunneling across the diode creates truly random signal, equivalent to thermal noise.
Sure, if you don't mind highly-exploitable biases in your random numbers. Randomness is not enough. You also need to obtain a uniform distribution. Simple hardware designs may also be influenced by external signals, which may be predictable or even controllable by the attacker. The fact that there is some quantum-derived randomness doesn't mean there is no predictability.
Doing this stuff right is hard. /dev/random in the Linux kernel actually does a fine job, better than any cost-effective hardware implementation will. Though, a cheap and biased hardware RNG can be used to help feed the kernel's entropy pool.
So, I would assume if there were other digital wallet type things on Android, they would be subject to the exact same vulnerability.
Unless they open an SSL/TLS connection prior to generating random numbers, or if they aren't dependent on randomness for security (unlikely).
IMO it also depends heavily on the nature of the course.
I really suck at memorizing anything, whether I write it down or not. Though writing it helps, assuming the process of writing doesn't cause me to miss something else I should have learned. That's basically why I ended up with a math degree, because in math courses (especially upper division) there is basically no point in trying to memorize anything. Either you understand it or you've got nothing, and what little you might have to memorize is all in the book anyway, in bolded and bulleted definitions and theorems. In those classes it's clearly far better to focus on listening, thinking and understanding; taking notes is a waste of time and a pure distraction.
In classes which are all about absorbing a big pile of facts, memorization aids are useful. In classes which are about understanding ideas, memorization is useless and so are notes. Courses which mix facts and ideas need a mixed approach, and are often the most difficult.
The best example I've seen of the latter is a freshman-level physics (mechanics, mostly) class my wife took, which didn't assume, nor explicate, any calculus. That class was a cast-iron bitch, because if you don't understand simple calculus you have to memorize a whole bunch of variations of every fundamental equation. For her final she had literally pages of equations to memorize, with no obvious relationship between expressions which are actually deeply related, plus she had to understand what all of them meant. The result was a class that was an order of magnitude more difficult than the same class with a calculus pre-req. I tried to get her to let me teach her the necessary calculus, promising that it wasn't more stuff to learn, but less. She didn't believe me and just gutted the course out as presented.
In a few years, perhaps technology will allow us to flip the problem. Rather than entering everything into the EMR and printing it out to have paper, you might be able to use paper as the data entry tool. If you could easily scan the paper and automatically extract the details needed to update the EMR, then you could have both. Of course, accurate automated analysis of doctors' handwriting is just a wee bit challenging. But machines should ultimately be able to do it as well as people, including being able to ask when they just can't make it out.
When those workers get to be about 35 YO, they'll be back to reality when they're looking for work and a place to live that they can afford.
Enjoy it while you can - your ass will be kicked to the curb before you know it.
Not at Google. Google employs lots of older engineers. I was hired at age 41 and I'm far from the oldest around. One of the guys I work with is in his 60s. The company still tends to be skewed towards the younger end of the demographics, but I'll bet the median age is in the low to mid-30s and rising. The Colorado office, where I work, tends to be older still, because the cost of living in the bay area drives away people with families.
Bear in mind that most Google employees are not "techies". They're sales reps selling ads. When you think Google, think "Mad Men", not rocket science.
Um, roughly 50% of Google's 40,000 full-time employees are software engineers. The other half make up everything else. I don't know numbers for sales reps, but I'd be surprised if they comprised 20% of the FTEs.
Oh, it's also worth pointing out that Wickard v. Fillburn was a direct result of FDR's court packing proposal. Like most of the related rulings of the time, it's bad law. Not that I see any realistic chance of reversing it any time soon.
Yes, I understand that's the position the courts have taken. The result is to destroy the notion that the constitution limits the scope of the federal government in any way, since absolutely anything can potentially affect commerce.
This is utter nonsense - the US government is allowed to regulate anyone conducting financial transactions within the US.
You mean interstate financial transactions, of course.
However, this article isn't about the US government, it's about New York state, which really does have legal authority to regulate any financial transactions in New York.
Exactly.
Don't forget the federal ban on marijuana possession and distribution.
I call them Obushma, because for the life of me I can't see any significant difference between them.