Oh, that's been figured out. What you do is get 40 random strangers to front you 50 cents each, with an offer to pay them back 1/40th of any payments that come in at 4%. Then you only loan your friend $18 and keep $2 for yourself. While you're at it you give $18 more to every homeless person you can find, and there isn't any way that anybody could lose money on such a deal. But, if they're worried they can all promise to pay each other $500 for each homeless guy who defaults.
So, my examples were meant to be illustrative - not suggestions that they specifically will be the wave of the future.
Maybe in the future everybody will insist on some kind of HDR light correction that requires a bazillion CPU cycles. Maybe the default scene transition fad of the day will become a morph rather than a wipe. My point is that anytime hardware takes off, we manage to think of something (perhaps frivolous) to keep it busy...
Yup, I see a lot of talk about tablets taking over the business, but I don't see it actually happening. I can see it in roles like sales and management. These are people who primary consume and present information, and generally don't capture much of it.
On the other hand, the people creating all the content that these people are displaying on their tablets will probably do it on a desktop (or a docked laptop).
I'm not sure about that - not until CPU becomes so trivial that a could server can do whatever you can possibly conceive of for virtually no cost.
Suppose I want to render a movie. That takes a boatload of CPU. Now, those CPUs can be in EC2, or they could be sitting in your office. However, you still need x cores running at y MIPS for z seconds.
For somebody who puts a movie together once a year the cloud model makes sense. Why buy a high-end workstation for something you don't do often. Instead you just pay some company a dollar or two to render your movie, and they can of course do it in parallel much faster.
On the other hand, if you're a videographer, and you render movies all day long, then paying somebody a few bucks for every 15 minutes of video you run through (and every time you make a correction) could get pretty expensive compared to just owning a computer.
Now, it is true that every two years it takes half as much CPU to do the same work. However, expectations constantly rise so the videographer working in 480p last year is working in 1080p next year, and maybe they'll be doing stereo video 10 years from now.
What the PC has going for it is about the highest performance and storage density per dollar you can get. It compromises in a lot of other areas, but there are a lot of consumer-oriented applications where this works well.
Yup - I'd go a step further and suggest that vulernability is pretty likely to be closely related to income.
In a society where people are paid in handfuls of rice per day, if you have a complicated problem you just requisition an extra sack of rice.
In a society like the US where even janitors are unionized and making $50k/yr, you automate the living daylights out of everything.
I saw an article about a military effort to build a highly automated destroyer - one that would operate with maybe a dozen men on board. Such a ship would probably work fine up until the point that something breaks. As soon as somebody so much as rakes it with machine gun fire you have a crew that is at 100% capacity just operating the device, and there is nobody around to start splicing cables.
When your ship has 300 men onboard you can recover even from modest torpedo hits.
Now, if the automation allows you to field 100x as many ships then maybe you can argue that even if half of them get sunk you still have more remaining capability and fewer dead sailors than you would from a single torpedo hit on a conventional ship. However, usually building 100x as many ships is not what is envisioned, but rather saving a few million dollars in operational costs (on a billion dollar ship).
Yup - I upgraded my desktop for $300, and the previous one cost me about $300 and lasted me around 4-5 years. There is no way I could have gotten that kind of performance out of a laptop. Oh, and in the process I migrated all my data from old hard drives to new ones, and did most of that while the desktop was fully operational (running various daemons from mythtv to samba).
Oh, and I'm using the same monitor I was using in around 1999 or so. I might just upgrade that to recover the electricity savings if I can find a decent cheap replacement, but I'm not throwing out an LCD panel every time I upgrade.
Desktops are always going to be the best bang for the buck.
My other concern would be that it could lead to wasteful peripherals.
Imagine a USB flash drive that pulls 5W, or an optical mouse that pulls 15W? If you remove the constraint on manufacturers to get the job done in x mW then unless people start buying devices based on power rating there will be a huge upwards drift, and I can look forward to increases in my electricity bill (with a double-hit from my cooling bill).
People don't even look at the speed ratings on flash drives today, let alone power draw...
Yup, and I suspect that most people running the browser on linux are using the chromium version, unless they downloaded it directly from Google. I imagine that most distros would prefer to start from the source, rather than try to support whatever version of libfoo Google chose to link against.
I've got chromium on Gentoo, and Chrome OS on a CR-48, and the only browser-related differences I can see are the ones listed above.
The sole objective of publicly traded corporations is whatever the stockholders want it to be.
Well, I'd go a step forward and make that: the sole objective of publicly traded corporations is whatever the executive management team can get away with.
Very few decisions "made by corporations" end up being in the interests of stockholders. The CEO just needs to convince the board that they are. Stockholders aren't watching what is going on 24x7. And, companies don't make decisions - executives do.
In a word, deregulation causes this sort of problem.
The old model was cost-plus. The utility creates a plan for how they'd like to operate, and how much electricity should cost to pay for it. The Public Utilities Commission would approve the plan. Then everybody follows the plan and you get charged the state-approved rate for your electricity use.
Typically these plans would include some level of redundancy/protection/etc so that the whole country wouldn't be in the dark ages if a tree shorted out one transmission line in the wrong place. That was factored into the expense, and the rates, that were charged.
Deregulation uses a different model. In deregulation generators get paid to sell electricity. If you generate electricity, and you don't sell it, then you don't get paid. Nobody approves any plans. In theory you have incentive to keep your plant running, since if you break down you aren't generating electricity, and thus you aren't getting paid. However, this breaks down in the area of redundancy and risk-prevention.
Redundancy by its definition means being able to produce more power than you can sell. That is just a loss to a generator. It has huge benefits for society, but the generators don't get paid for that, so they have no incentive to provide it. To some extent you can argue that they stand to make profit if a competitor goes down, but the economics of that usually don't work out.
Let's look at it from a supply vs consumer standpoint. I run a hospital. My electric bill is $10k/month. A one-day electrical interruption of any kind probably would easily cost me $50k to deal with. Now, suppose I run a generator. I make $10k/month selling power to a hospital. If my plant goes down for a day and that is the straw that breaks the camel's back taking out half of the East Coast of the USA, I lose about $300 ($10k/30) in revenue selling electricity to that hospital. Now, that multiples across many more customers, but at the end of the day I have a lot less incentive to produce power than my consumers have to get it reliably.
In order to fix this you need to either get rid of deregulation, or you need to create a framework that provides incentive to have generation on standby or just mandate that companies can't sell more than 80% of their power so that the level playing field just increases the market price a little.
I read an article about this problem back in the 80s, actually. At that time the concern was that a well-organized terrorist (or foreign - remember this was the cold war) attack could take out the power grid for a long time.
A typical city does not have a lot of redundancy in its substation network - it can lose one or two, but not more than that. And, when power goes out it is usually due to breakers tripping or things like that - not deliberate attack.
The threat model in the article was a bunch of infiltrators with rifles that at a coordinated time start shooting at insulators in substations. This would destroy the equipment, and there is not enough of a supply of these transformers in stock to replace them quickly. Plus, if this was coordinated nationwide you could literally end up with 90% of the USA having no electricity for six months, which certainly complicates building new transformers and getting them where they need to be.
If your data contains any personal information about another person, you are placing the privacy of that person in the hands of an organisation you do not control, and upon whom you cannot enforce any legislative restrictions.
Yes, but it does give you somebody to point your finger at when there is a data loss. That is almost always enough for a manager to get to keep his job (assuming he wasn't already promoted for all the cost-savings years ago). Often it is enough to keep a company out of legal trouble (but this is AMAZON - what person wouldn't assume they know what they're doing with personal data?).
Unless a company is privately owned there is rarely enough oversight to keep it from betting the farm on the latest get-rich-quick scheme. Just look at Wall Street a few years ago.
Yup, even T-Mobile has new "unlimited" plans that throttle at some level. The only difference between those plans and the grandfathered ones is that the new ones tell you what the threshold is. They always throttled you to 2G if you used too much bandwidth...
I think it is the cable company mentality. Back in the 80s cable companies were all upset that some of their consumers would buy one cable box, and then split the coax to their TV and run it to a second TV. This of course was a serious inconvenience since you'd have to change rooms to change channels, but cable company executives probably spent a small fortune in meeting time trying to figure out ways to prevent it. I'm surprised they didn't lobby for the right to inspect homes at random, and the contracts invariably would say one TV per tuner box. In the end the behavior is very hard to police.
Cable companies had the same concerns about people being able to hook up more than one computer to a router. This actually is detectable most of the time right now, and they had a general resentment of not being able to charge for every computer, iPod, Tivo, and whatever else might be on that network. They never did end up policing this since it would have invariably led to consumer backlash.
Look at DRM where the goal is to get you to keep buying the same content over and over, eliminate libraries, and get rid of people loaning stuff to friends.
This is just another example of the corporate entitlement mentality. Why innovate and sell a new or better product when you can instead try to force people (through oligopoly power) to keep buying the same product over and over?
I'd love to see Cyanogenmod have better VPN support, whether I use tether or not.
Right now you can connect to a VPN, but only traffic for hosts on the VPN is transferred over the VPN.
I'd like an arrangement where 100% of all my phone's traffic (controlled at the OS level) is sent through the VPN. My home broadband can easily handle the data a phone will generate, and I'm not concerned with a little extra latency with what I do.
There is an openvpn setting that is supposed to do something like this, but it apparently even sends DHCP traffic through the VPN which means you'll lose your lease (and it is fussy to set up anyway). Plus, if the VPN link drops it will just send traffic over the normal network interface, which means you lose privacy on whatever you were sending.
I would think from an implementation standpoint all it needs to do is: 1. Create a VPN tunnel on a network interface (all VPN software does this already). 2. Make the VPN tunnel the default route (just a routing table change). 3. Use netfilter/etc to block all but carefully-selected traffic to the phone's real ethernet interface. The VPN service obviously needs access, as does the DHCP service. 4. Ensure that there are no leaks if the VPN goes down for a minute (user-controllable setting) - making sure the firewall rules don't get torn down if the VPN service hiccups is all that should require.
Then all the provider gets is a big stream of VPN traffic and some DHCP packets. No way to steal cookies, insert ads, track behavior, or violate net-neutrality. They can charge by the GB and they can count how many GB I use, so there is no theft of service.
Keep in mind that some of the largest areas of spending are entitlements, and those involve issuing bazillions of medicare payments for doctor visits and surgeries, and a modest check every month to millions of people. In a sort by amount of individual checks, they'd be very low on the list compared to buying buildings or tanks, but they add up to quite a bit. Actually, I'm not sure where the military would even rank on something like this - it isn't like the government just goes into the Lockheed store and buys 10 F-22s - there are probably contracts involving regular payments and such.
In the aggregate the largest areas of spending are social security, medicare, and the military. When you talk about future obligation then the military actually drops off the picture - it is a lot easier to stop buying bombers than to tell somebody that despite paying social security taxes their entire life they're not going to collect as much as their parents did.
So, at this point in most conversations everybody starts going back and forth about which of these three is a legitimate and important function of government, and in the end since we can't agree we keep paying for them all.
Well, where the behavior of nations is concerned, the only thing that regulates who can seize what is the size of everybody's armies. The US army being what it is, seizing property isn't going to be practical for any individual nation - at least, not at present..
I think some of this is also just the increasing education of the US citizen.
The reality is that companies have strategically defaulted on loans all the time. You take a credit rating hit, and you might get collected on if you have anything to collect. However, it is just a business decision.
I think that in the past loans used to be seen more as a MORAL obligation, and not merely a financial arrangement. There was something inherently wrong about not paying back a loan, and an element of shame in doing it. Just as a normal person wouldn't make a cold calculation weighing the opportunity afforded by stabbing somebody against the risk of capture and likely prison term, a normal person back then wouldn't just look at the numbers and say "oh well, let's just borrow a ton of cash, have fun, and default."
Today the average person is a lot closer to how companies operate and sees this sort of thing happen all the time among the rich and in business. If loans are nothing more than a financial instrument, and a choice to default is nothing more than a decision that incurs certain short-term and long-term benefits and payments, then the decision to default is a much easier one to make - especially with laws being what they are.
NPR did a good episode on this not long ago. If you're some 70-year-old who bought an investment property for $300k and it is now worth $200k, and you can inexpensively stall foreclosure for a year or two while still collecting rent, then why not just stop making your payments entirely and let the bank worry about the $100k loss? Chances are you'll never have to buy another house, and you can make quite a bit of money while the legal system grinds.
Ok, so it is like the guy who makes $50k/yr with a $350k mortgage, and is applying for a car loan whose PAYMENTS are $100k/yr. Clearly, that is a much more solid financial position to be in.
Anybody with half a brain could see at least 2-3 things in the sentence above that should cause any investor to run to the door FAST.
Yup, that's the problem with value-by-fiat: it breaks down at your jurisdictional lines. Just look at the California energy crisis.
Market rate for power was $x. Consumers paid $y per law. Utilities start losing tons of money. State steps in and tries to force generators to accept $y. Within the US they were able to twist arms to an extent (at least forcing them to deliver on credit if nothing else). However, when they needed power from north of the border, the Canadian utilities demanded cash in advance and were all too willing to flip off the switch.
People aren't idiots - if you can't effectively hold a gun to their heads, then you can't use legislation to change their actions.
All those dollar bills in your wallet are only as valuable as we collectively agree they are.
Well then, let's collectively agree they are worth a mansion and a yacht for each citizen, shall we ? As it turns out, the long term value of money is not the result of psychology, rather a simple division of the economic output of the country by the outstanding monetary base. If some guy in Washington keeps printing those bills, they will eventually drop to zero, no collective agreement needed.
Well, it is a little of both - like any measure of value.
To the guy wanting to buy a house, $100 should buy a house. To the guy wanting to build a house, $100 should buy a bucket of nails. The two meet somewhere in the middle, and that is the value of money.
In practice, your description is a good way to get a sense of where people will end up. However, how do you measure economic output without using dollars? If you measure economic output in dollars, then your definition of the value of the dollar is circular.
If all people did was make nails, and trade them for dollars, then the value of a dollar in nails would be trivially measured as you stated. However, the whole point of currency is allowing us to value financial transactions that are not obviously comparable (such as how many buckets of nails to pay your doctor for an operation, when he has no need for nails).
I wonder if this is a technology or a human issue.
Humans simply aren't good at staring at a sea of harmless stuff and then spotting something dangerous.
I remember reading about a company that came up with an x-ray machine that had a game of sorts built into it. The scanner would periodically put a picture of a gun or a knife or whatever in the actual images of baggage, and the operator had to push a button to remove it. If the operator missed something then it would alert the supervisor to change out the operator. This gives the operator something to do so that they don't zone out. Obviously if you push the button and the gun doesn't go away then you have a problem to deal with.
Finally - this thing talks about mm waves - but I thought that the machines in the US were backscatter x-rays? The wavelength of an X-ray is WAY smaller than a mm. I could see how mm-wavelength radiation is going to be more easily fooled by wet clothes/etc.
Absolutely true on all points, and I still stand by what I said.
What I was getting at is that more type 2 diabetics should be treated with pumps. I think that too many doctors settle for "the best that can be done with oral meds."
If that is good glucose control, then fine. However, often it isn't. From what I've both read and seen firsthand unless you're talking about an 85-year-old who you're consigned to putting in "hospice," you're much better off being more aggressive. While shots or a pump are inconvenient, the kinds of health problems that come up after 5-10 years of high blood sugar are FAR worse.
By all means try to fix the insulin resistance first. But, until you've figured out how to fix that pumping a patient full of insulin will at least keep them from having to get their more essential arteries stented, their eyes lasered, and so on. The best way to do that I've seen is aggressive monitoring with frequent medication tweaks until they're stable, and if that isn't working a pump gives you the most fine-grained control possible.
Slashdot Mirror
Oh, that's been figured out. What you do is get 40 random strangers to front you 50 cents each, with an offer to pay them back 1/40th of any payments that come in at 4%. Then you only loan your friend $18 and keep $2 for yourself. While you're at it you give $18 more to every homeless person you can find, and there isn't any way that anybody could lose money on such a deal. But, if they're worried they can all promise to pay each other $500 for each homeless guy who defaults.
So, my examples were meant to be illustrative - not suggestions that they specifically will be the wave of the future.
Maybe in the future everybody will insist on some kind of HDR light correction that requires a bazillion CPU cycles. Maybe the default scene transition fad of the day will become a morph rather than a wipe. My point is that anytime hardware takes off, we manage to think of something (perhaps frivolous) to keep it busy...
Yup, I see a lot of talk about tablets taking over the business, but I don't see it actually happening. I can see it in roles like sales and management. These are people who primary consume and present information, and generally don't capture much of it.
On the other hand, the people creating all the content that these people are displaying on their tablets will probably do it on a desktop (or a docked laptop).
I'm not sure about that - not until CPU becomes so trivial that a could server can do whatever you can possibly conceive of for virtually no cost.
Suppose I want to render a movie. That takes a boatload of CPU. Now, those CPUs can be in EC2, or they could be sitting in your office. However, you still need x cores running at y MIPS for z seconds.
For somebody who puts a movie together once a year the cloud model makes sense. Why buy a high-end workstation for something you don't do often. Instead you just pay some company a dollar or two to render your movie, and they can of course do it in parallel much faster.
On the other hand, if you're a videographer, and you render movies all day long, then paying somebody a few bucks for every 15 minutes of video you run through (and every time you make a correction) could get pretty expensive compared to just owning a computer.
Now, it is true that every two years it takes half as much CPU to do the same work. However, expectations constantly rise so the videographer working in 480p last year is working in 1080p next year, and maybe they'll be doing stereo video 10 years from now.
What the PC has going for it is about the highest performance and storage density per dollar you can get. It compromises in a lot of other areas, but there are a lot of consumer-oriented applications where this works well.
Yup - I'd go a step further and suggest that vulernability is pretty likely to be closely related to income.
In a society where people are paid in handfuls of rice per day, if you have a complicated problem you just requisition an extra sack of rice.
In a society like the US where even janitors are unionized and making $50k/yr, you automate the living daylights out of everything.
I saw an article about a military effort to build a highly automated destroyer - one that would operate with maybe a dozen men on board. Such a ship would probably work fine up until the point that something breaks. As soon as somebody so much as rakes it with machine gun fire you have a crew that is at 100% capacity just operating the device, and there is nobody around to start splicing cables.
When your ship has 300 men onboard you can recover even from modest torpedo hits.
Now, if the automation allows you to field 100x as many ships then maybe you can argue that even if half of them get sunk you still have more remaining capability and fewer dead sailors than you would from a single torpedo hit on a conventional ship. However, usually building 100x as many ships is not what is envisioned, but rather saving a few million dollars in operational costs (on a billion dollar ship).
Yup - I upgraded my desktop for $300, and the previous one cost me about $300 and lasted me around 4-5 years. There is no way I could have gotten that kind of performance out of a laptop. Oh, and in the process I migrated all my data from old hard drives to new ones, and did most of that while the desktop was fully operational (running various daemons from mythtv to samba).
Oh, and I'm using the same monitor I was using in around 1999 or so. I might just upgrade that to recover the electricity savings if I can find a decent cheap replacement, but I'm not throwing out an LCD panel every time I upgrade.
Desktops are always going to be the best bang for the buck.
My other concern would be that it could lead to wasteful peripherals.
Imagine a USB flash drive that pulls 5W, or an optical mouse that pulls 15W? If you remove the constraint on manufacturers to get the job done in x mW then unless people start buying devices based on power rating there will be a huge upwards drift, and I can look forward to increases in my electricity bill (with a double-hit from my cooling bill).
People don't even look at the speed ratings on flash drives today, let alone power draw...
Yup, and I suspect that most people running the browser on linux are using the chromium version, unless they downloaded it directly from Google. I imagine that most distros would prefer to start from the source, rather than try to support whatever version of libfoo Google chose to link against.
I've got chromium on Gentoo, and Chrome OS on a CR-48, and the only browser-related differences I can see are the ones listed above.
Hmm, even a little more complicated than this, since I only want VPN traffic to go to that address, not non-VPN traffic to the same IP.
Depends - if the exploit works on android phones then I'd expect the patch to be deployed in anywhere from six months to never...
The sole objective of publicly traded corporations is whatever the stockholders want it to be.
Well, I'd go a step forward and make that: the sole objective of publicly traded corporations is whatever the executive management team can get away with.
Very few decisions "made by corporations" end up being in the interests of stockholders. The CEO just needs to convince the board that they are. Stockholders aren't watching what is going on 24x7. And, companies don't make decisions - executives do.
In a word, deregulation causes this sort of problem.
The old model was cost-plus. The utility creates a plan for how they'd like to operate, and how much electricity should cost to pay for it. The Public Utilities Commission would approve the plan. Then everybody follows the plan and you get charged the state-approved rate for your electricity use.
Typically these plans would include some level of redundancy/protection/etc so that the whole country wouldn't be in the dark ages if a tree shorted out one transmission line in the wrong place. That was factored into the expense, and the rates, that were charged.
Deregulation uses a different model. In deregulation generators get paid to sell electricity. If you generate electricity, and you don't sell it, then you don't get paid. Nobody approves any plans. In theory you have incentive to keep your plant running, since if you break down you aren't generating electricity, and thus you aren't getting paid. However, this breaks down in the area of redundancy and risk-prevention.
Redundancy by its definition means being able to produce more power than you can sell. That is just a loss to a generator. It has huge benefits for society, but the generators don't get paid for that, so they have no incentive to provide it. To some extent you can argue that they stand to make profit if a competitor goes down, but the economics of that usually don't work out.
Let's look at it from a supply vs consumer standpoint. I run a hospital. My electric bill is $10k/month. A one-day electrical interruption of any kind probably would easily cost me $50k to deal with. Now, suppose I run a generator. I make $10k/month selling power to a hospital. If my plant goes down for a day and that is the straw that breaks the camel's back taking out half of the East Coast of the USA, I lose about $300 ($10k/30) in revenue selling electricity to that hospital. Now, that multiples across many more customers, but at the end of the day I have a lot less incentive to produce power than my consumers have to get it reliably.
In order to fix this you need to either get rid of deregulation, or you need to create a framework that provides incentive to have generation on standby or just mandate that companies can't sell more than 80% of their power so that the level playing field just increases the market price a little.
I read an article about this problem back in the 80s, actually. At that time the concern was that a well-organized terrorist (or foreign - remember this was the cold war) attack could take out the power grid for a long time.
A typical city does not have a lot of redundancy in its substation network - it can lose one or two, but not more than that. And, when power goes out it is usually due to breakers tripping or things like that - not deliberate attack.
The threat model in the article was a bunch of infiltrators with rifles that at a coordinated time start shooting at insulators in substations. This would destroy the equipment, and there is not enough of a supply of these transformers in stock to replace them quickly. Plus, if this was coordinated nationwide you could literally end up with 90% of the USA having no electricity for six months, which certainly complicates building new transformers and getting them where they need to be.
If your data contains any personal information about another person, you are placing the privacy of that person in the hands of an organisation you do not control, and upon whom you cannot enforce any legislative restrictions.
Yes, but it does give you somebody to point your finger at when there is a data loss. That is almost always enough for a manager to get to keep his job (assuming he wasn't already promoted for all the cost-savings years ago). Often it is enough to keep a company out of legal trouble (but this is AMAZON - what person wouldn't assume they know what they're doing with personal data?).
Unless a company is privately owned there is rarely enough oversight to keep it from betting the farm on the latest get-rich-quick scheme. Just look at Wall Street a few years ago.
Yup, even T-Mobile has new "unlimited" plans that throttle at some level. The only difference between those plans and the grandfathered ones is that the new ones tell you what the threshold is. They always throttled you to 2G if you used too much bandwidth...
I think it is the cable company mentality. Back in the 80s cable companies were all upset that some of their consumers would buy one cable box, and then split the coax to their TV and run it to a second TV. This of course was a serious inconvenience since you'd have to change rooms to change channels, but cable company executives probably spent a small fortune in meeting time trying to figure out ways to prevent it. I'm surprised they didn't lobby for the right to inspect homes at random, and the contracts invariably would say one TV per tuner box. In the end the behavior is very hard to police.
Cable companies had the same concerns about people being able to hook up more than one computer to a router. This actually is detectable most of the time right now, and they had a general resentment of not being able to charge for every computer, iPod, Tivo, and whatever else might be on that network. They never did end up policing this since it would have invariably led to consumer backlash.
Look at DRM where the goal is to get you to keep buying the same content over and over, eliminate libraries, and get rid of people loaning stuff to friends.
This is just another example of the corporate entitlement mentality. Why innovate and sell a new or better product when you can instead try to force people (through oligopoly power) to keep buying the same product over and over?
I'd love to see Cyanogenmod have better VPN support, whether I use tether or not.
Right now you can connect to a VPN, but only traffic for hosts on the VPN is transferred over the VPN.
I'd like an arrangement where 100% of all my phone's traffic (controlled at the OS level) is sent through the VPN. My home broadband can easily handle the data a phone will generate, and I'm not concerned with a little extra latency with what I do.
There is an openvpn setting that is supposed to do something like this, but it apparently even sends DHCP traffic through the VPN which means you'll lose your lease (and it is fussy to set up anyway). Plus, if the VPN link drops it will just send traffic over the normal network interface, which means you lose privacy on whatever you were sending.
I would think from an implementation standpoint all it needs to do is:
1. Create a VPN tunnel on a network interface (all VPN software does this already).
2. Make the VPN tunnel the default route (just a routing table change).
3. Use netfilter/etc to block all but carefully-selected traffic to the phone's real ethernet interface. The VPN service obviously needs access, as does the DHCP service.
4. Ensure that there are no leaks if the VPN goes down for a minute (user-controllable setting) - making sure the firewall rules don't get torn down if the VPN service hiccups is all that should require.
Then all the provider gets is a big stream of VPN traffic and some DHCP packets. No way to steal cookies, insert ads, track behavior, or violate net-neutrality. They can charge by the GB and they can count how many GB I use, so there is no theft of service.
Keep in mind that some of the largest areas of spending are entitlements, and those involve issuing bazillions of medicare payments for doctor visits and surgeries, and a modest check every month to millions of people. In a sort by amount of individual checks, they'd be very low on the list compared to buying buildings or tanks, but they add up to quite a bit. Actually, I'm not sure where the military would even rank on something like this - it isn't like the government just goes into the Lockheed store and buys 10 F-22s - there are probably contracts involving regular payments and such.
In the aggregate the largest areas of spending are social security, medicare, and the military. When you talk about future obligation then the military actually drops off the picture - it is a lot easier to stop buying bombers than to tell somebody that despite paying social security taxes their entire life they're not going to collect as much as their parents did.
So, at this point in most conversations everybody starts going back and forth about which of these three is a legitimate and important function of government, and in the end since we can't agree we keep paying for them all.
Well, where the behavior of nations is concerned, the only thing that regulates who can seize what is the size of everybody's armies. The US army being what it is, seizing property isn't going to be practical for any individual nation - at least, not at present..
I think some of this is also just the increasing education of the US citizen.
The reality is that companies have strategically defaulted on loans all the time. You take a credit rating hit, and you might get collected on if you have anything to collect. However, it is just a business decision.
I think that in the past loans used to be seen more as a MORAL obligation, and not merely a financial arrangement. There was something inherently wrong about not paying back a loan, and an element of shame in doing it. Just as a normal person wouldn't make a cold calculation weighing the opportunity afforded by stabbing somebody against the risk of capture and likely prison term, a normal person back then wouldn't just look at the numbers and say "oh well, let's just borrow a ton of cash, have fun, and default."
Today the average person is a lot closer to how companies operate and sees this sort of thing happen all the time among the rich and in business. If loans are nothing more than a financial instrument, and a choice to default is nothing more than a decision that incurs certain short-term and long-term benefits and payments, then the decision to default is a much easier one to make - especially with laws being what they are.
NPR did a good episode on this not long ago. If you're some 70-year-old who bought an investment property for $300k and it is now worth $200k, and you can inexpensively stall foreclosure for a year or two while still collecting rent, then why not just stop making your payments entirely and let the bank worry about the $100k loss? Chances are you'll never have to buy another house, and you can make quite a bit of money while the legal system grinds.
Ok, so it is like the guy who makes $50k/yr with a $350k mortgage, and is applying for a car loan whose PAYMENTS are $100k/yr. Clearly, that is a much more solid financial position to be in.
Anybody with half a brain could see at least 2-3 things in the sentence above that should cause any investor to run to the door FAST.
Yup, that's the problem with value-by-fiat: it breaks down at your jurisdictional lines. Just look at the California energy crisis.
Market rate for power was $x. Consumers paid $y per law. Utilities start losing tons of money. State steps in and tries to force generators to accept $y. Within the US they were able to twist arms to an extent (at least forcing them to deliver on credit if nothing else). However, when they needed power from north of the border, the Canadian utilities demanded cash in advance and were all too willing to flip off the switch.
People aren't idiots - if you can't effectively hold a gun to their heads, then you can't use legislation to change their actions.
All those dollar bills in your wallet are only as valuable as we collectively agree they are.
Well then, let's collectively agree they are worth a mansion and a yacht for each citizen, shall we ? As it turns out, the long term value of money is not the result of psychology, rather a simple division of the economic output of the country by the outstanding monetary base. If some guy in Washington keeps printing those bills, they will eventually drop to zero, no collective agreement needed.
Well, it is a little of both - like any measure of value.
To the guy wanting to buy a house, $100 should buy a house. To the guy wanting to build a house, $100 should buy a bucket of nails. The two meet somewhere in the middle, and that is the value of money.
In practice, your description is a good way to get a sense of where people will end up. However, how do you measure economic output without using dollars? If you measure economic output in dollars, then your definition of the value of the dollar is circular.
If all people did was make nails, and trade them for dollars, then the value of a dollar in nails would be trivially measured as you stated. However, the whole point of currency is allowing us to value financial transactions that are not obviously comparable (such as how many buckets of nails to pay your doctor for an operation, when he has no need for nails).
I wonder if this is a technology or a human issue.
Humans simply aren't good at staring at a sea of harmless stuff and then spotting something dangerous.
I remember reading about a company that came up with an x-ray machine that had a game of sorts built into it. The scanner would periodically put a picture of a gun or a knife or whatever in the actual images of baggage, and the operator had to push a button to remove it. If the operator missed something then it would alert the supervisor to change out the operator. This gives the operator something to do so that they don't zone out. Obviously if you push the button and the gun doesn't go away then you have a problem to deal with.
Finally - this thing talks about mm waves - but I thought that the machines in the US were backscatter x-rays? The wavelength of an X-ray is WAY smaller than a mm. I could see how mm-wavelength radiation is going to be more easily fooled by wet clothes/etc.
Absolutely true on all points, and I still stand by what I said.
What I was getting at is that more type 2 diabetics should be treated with pumps. I think that too many doctors settle for "the best that can be done with oral meds."
If that is good glucose control, then fine. However, often it isn't. From what I've both read and seen firsthand unless you're talking about an 85-year-old who you're consigned to putting in "hospice," you're much better off being more aggressive. While shots or a pump are inconvenient, the kinds of health problems that come up after 5-10 years of high blood sugar are FAR worse.
By all means try to fix the insulin resistance first. But, until you've figured out how to fix that pumping a patient full of insulin will at least keep them from having to get their more essential arteries stented, their eyes lasered, and so on. The best way to do that I've seen is aggressive monitoring with frequent medication tweaks until they're stable, and if that isn't working a pump gives you the most fine-grained control possible.