God, please don't tell me you're that naive. "It's not easily accessible to the government"?? It's one simple subpoena away, ffs!
You misunderstood me. I was talking about warrantless wiretaps.
That's the *last* thing you want if your goal is to circumvent government surveillance.
Again, I was not talking about "circumventing government surveillance" - to do that you need encryption, steganography, some covert channel - that kind of shit. But I was talking about a simple way to avoid wholesale monitoring by government. Fishing expeditions if you will.
The point is to be reasonable, not paranoid. I don't want the government to read my emails when I tell my buddies that they must go watch "Hangover" this weekend, but I will not jump through security hoops to deliver that message.
What happens is that they (NSA) DoS-ing investigative resources. FBI and such have only so many men in the field to check the facts. As a result, the ability (probability) to identify true threats goes way down.
The same goes for other after 9/11 security "improvements" like, for example, indiscriminate "deep background investigation" of immigrants - the queue became so long, that it takes years now (not shitting) to get men from "interesting" countries checked! And I'm not talking about nutcases holed up somewhere in Pakistan mountains - I'm talking about people who already walk the streets of the US!
Too bad we don't know how to imitate free market's ability to optimally allocate resources in rigid government setups...
The problem is - almost nobody uses this. So you can be singled out on that fact alone.
As a poor man's solution, one can use Gmail over https (they have that option now): in my case all my friends have gmail accounts. It's not easily accessible to the government (assuming google's internal traffic is not tapped).
This of course exposes you to Google, but at least there is a good chance, that it's not subject to warrantless wiretaps.
On top of that you can encrypt so google is off too.
ABC(7) and CBS(9) turned off analog AND moved digital from UHF to VHF. As result, huge number of people lost these channels because of UHF-only antennas they use. You can "rescan" as many times as you want - if there is not much signal to begin with...
The thing is - people were watching digital already, and were not aware of the frequency move - the stations didn't warn anybody...
On top of everything it seems that digital does not go that far riding on VHF comparing to UHF (at least at the power level they transmitting)
Just a small point: in many places police officers are forbidden to follow at high speed suspects of non-violent crimes.
That's because of large number of deaths among bystanders.
I think, your math is wrong. Not by itself though.
You see, his calculations are based on fuel costs (not on hard $$$). And those costs grow in accord with interest rates (you know fuel is that kind of commodity).
So his calculations in that regard are alright (more or less).
The thing is: the money is from "research fund" and not from "bailout fund".
And considering that Tesla is de facto leader in the field AND cash positive AND an american company - they are actually entitled to the money.
The gov't shouldn't loan money to start-ups unless it gets a share of the company... just like any other VC.
VCs in general do not loan the money - they get partial ownership of the company which (the share) they hope to sell to others later.
Basically if you loan - you expect to be repaid, if you own shares then company does not owe you any money - it gave you shares already. It's only when company goes into bankruptcy THEN creditors become owners.
With one exception for banks - there are different rules involved to simplify government takeover (for good reasons actually)
There is a reason that it does not exist in any place on earth. It's an abstraction, much like the opposite - communism.
You see - there is a problem with externalities (negative and positive).
But our problem is that when our government starts to regulate in places where regulation is needed (e.g. pollution control) it does not know when to stop.
Even worse: regulation (correctly made or not) in one area forces your hand in another.
In this case - because oil products get huge subsidies in form of tax going for all the military hardware in the Middle East (and note - not from gas taxes), you have to subsidize research into efficient transportation because market does not favor it.
Remember oil spike? Market was alright pushing huge sums of money into alternative energy research.
What about Metaram DDR3 modules? Hynix started to roll them out.
Basically they make largish modules using lower capacity (but much cheaper) chips, buffered in such way that to the system they look like slightly slower high capacity ones.
There was a news story few days ago, saying that Intel just certified them.
No, they are not. You forgetting added wight (need a bigger and, alas, heavier battery to accelerate), loss of cargo space, cost of maintenance, cost of gas, even weight of gas. It all adds up in a bad way.
You see, hybrids don't make much financial sense now (go see Toyota dealership's parking lot, after Prius hype is over). And with better battery technology they won't make sense tomorrow.
They are simply a temporary kludge on the way to pure electrics.
About "drastic" improvements: current battery technology is already good enough for second, commute only car. With gradual improvements (or BetterPlace like schemes) it will become a primary car.
"native Kindle books" are not tied to the Amazon account.
Kindle's native format is Mobipocket. There are free converters from pdf, html, word.
Simply convert, connect Kindle to your computer via usb and upload.
What people seems don't understand is that kindle and kindle like devices are mp3 players for books. And there are not only free classic books out there in the internet. If you know where to look you can find pretty much any book (e.g. pirated Potter books appear about 10 hours after release).
So Kindle is not an "attack on the first sale doctrine" but attack on the whole publishers racket.
3. Find me a computer that can use a RAM drive that can be persistent through reboots without having to save the contents to something else.
That's the thing - you don't need to. Your application should be able to use it naturally.
Just make sure you have a modern CPU which can handle it without performance degradation due to paging complications.
That's all in the game. This is the case with any risk measure, alone or in combination with others. The worst-case scenario is the Dies Irae; you cannot take it into account if you want to do any business at all.
Here's the thing: you cannot put it "into the game" when you use incorrect distribution (as they did with VaR). Otherwise pretty much everything (including the world end) can be taken into account given the right price (and possibility of payments in hard currency). The main problem will become the "gambler ruin" (in real "act of God" situations) which will drive premiums even higher than the distribution calls for. Plus, by all accounts, 2008 was not the year the world ended (only seemed that way).
So VaR as used by them was not even a proper "risk measure" - that was one of mistakes they did thinking that it was.
Sure, if I were working for a Wall Street bank and used a more realistic model I would get my ass fired in no time for what it would look like being so risk averse. But that's a systematic problem with businesses today.
The point was (and I am sorry if it didn't come across clearly), that though banks are better diversified on the micro level, this leads to them holding similar portfolios, introducing systemic risk. That is, banks, as an aggregate, are difersified poorly as a result (among others) of Basel-induced VaR contsraints.
Ok. So they diversified they asses to such extent that overall financial system "connectedness" grew high enough to make it unstable.
I smell a common mistake here: "rational players" and "lucky players" are indistinguishable at this point.
That's the problem with markets.
Hell, even with Buffet it is hard to be all one hundred percent sure that he's indeed a genius and not just one really-really lucky dude.
VaR is a pretty decent risk measure on a micro scale.
No, it's not. More precisely it is but only for 99.9% of days. That's Taleb's point. You look at the VaR for 3 years in a row and it's always right, you grow to trust it, and then the big whoop comes.
The real problem with it is that VaR constraints tend to make banks less diversified
Actually it's other way around. It's the property to diversify as much as you please that VaR provides. That's why improper calculation of risks involved caused such a wide spread of problems.
What the financial markets need is a better understanding of the causal links between risks and to price the returns on investments accordingly.
And that's the problem - what are you proposing can't be done. Even fully deterministic systems can display unpredictable behavior (e.g. 3 body problem). Chaotic systems with random events thrown in WILL display that "fat tail" thing even when you know ALL causal relationships in them.
Taleb made his money by not correctly estimating what's the real risk situation is, he made it by betting on his belief that others are off by a long shot.
If I remember correctly, Nikola Tesla invented an apparatus to wreck buildings from within by finding building's resonant frequency and pounding at it.
Kinda like modern mass dampers, but active and, well, not damping much...
Sure, but write leveling algorithms are essentially black boxes.
The thing is - to avoid trashing SSD when it's almost full you need to actually move data around - i.e. move "stable" data sectors from "fresh" cells into highly used ones. So the obvious problems are:
1. What if my application is such, that it will "chase" the same blocks that the controller targets for moving?
2. When do you do that background moving (reading&writing) thing? If I pound the disk continuously it will suddenly lose 50% write performance (which is not too hot to begin with).
So the numbers for performance and reliability are very application based (true - situation is the same as with mechanical drives but much less predictable). I.e. you don't know how well particular drive model will do until you actually try it (and for a long period of time)
So you select an SLC and sit with your fingers crossed.
1. You own a battery (and take good care of it).
2. When you go for a trip, first thing you do is stop by the local "U-bat" and install leased battery (your's going into their cool basement with complementary recharge)
3. You drive as far as you please, changing from one leased battery to another.
4. After returning home you stop by aforementioned "U-bat" to install your battery back.
If for some reason you skipped the step 2 the only downside is that there will be additional shipping charge.
So essentially, the only things that are needed: discharge meters on leased batteries and crypto lockup on your battery so that it cannot be used with any other car.
Slashdot Mirror
God, please don't tell me you're that naive. "It's not easily accessible to the government"?? It's one simple subpoena away, ffs!
You misunderstood me. I was talking about warrantless wiretaps.
That's the *last* thing you want if your goal is to circumvent government surveillance.
Again, I was not talking about "circumventing government surveillance" - to do that you need encryption, steganography, some covert channel - that kind of shit. But I was talking about a simple way to avoid wholesale monitoring by government. Fishing expeditions if you will.
The point is to be reasonable, not paranoid. I don't want the government to read my emails when I tell my buddies that they must go watch "Hangover" this weekend, but I will not jump through security hoops to deliver that message.
And not just "profiling".
What happens is that they (NSA) DoS-ing investigative resources. FBI and such have only so many men in the field to check the facts. As a result, the ability (probability) to identify true threats goes way down.
The same goes for other after 9/11 security "improvements" like, for example, indiscriminate "deep background investigation" of immigrants - the queue became so long, that it takes years now (not shitting) to get men from "interesting" countries checked! And I'm not talking about nutcases holed up somewhere in Pakistan mountains - I'm talking about people who already walk the streets of the US!
Too bad we don't know how to imitate free market's ability to optimally allocate resources in rigid government setups...
The problem is - almost nobody uses this. So you can be singled out on that fact alone.
As a poor man's solution, one can use Gmail over https (they have that option now): in my case all my friends have gmail accounts. It's not easily accessible to the government (assuming google's internal traffic is not tapped). This of course exposes you to Google, but at least there is a good chance, that it's not subject to warrantless wiretaps.
On top of that you can encrypt so google is off too.
That's a major problem in Washington DC area.
ABC(7) and CBS(9) turned off analog AND moved digital from UHF to VHF. As result, huge number of people lost these channels because of UHF-only antennas they use. You can "rescan" as many times as you want - if there is not much signal to begin with...
The thing is - people were watching digital already, and were not aware of the frequency move - the stations didn't warn anybody...
On top of everything it seems that digital does not go that far riding on VHF comparing to UHF (at least at the power level they transmitting)
I agree.
Just a small point: in many places police officers are forbidden to follow at high speed suspects of non-violent crimes. That's because of large number of deaths among bystanders.
You're lucky (in some sense).
If you were involved professionally in any kind of sensitive subject, you'd be fired, and be unemployable ever after.
Remember that poor (well, not anymore) schmuck in the white powder via USPS case?
I think, your math is wrong. Not by itself though.
You see, his calculations are based on fuel costs (not on hard $$$). And those costs grow in accord with interest rates (you know fuel is that kind of commodity).
So his calculations in that regard are alright (more or less).
The gov't shouldn't loan money to start-ups unless it gets a share of the company... just like any other VC.
VCs in general do not loan the money - they get partial ownership of the company which (the share) they hope to sell to others later. Basically if you loan - you expect to be repaid, if you own shares then company does not owe you any money - it gave you shares already. It's only when company goes into bankruptcy THEN creditors become owners.
With one exception for banks - there are different rules involved to simplify government takeover (for good reasons actually)
Capitalism would work if you let it.
No, it won't.
There is a reason that it does not exist in any place on earth. It's an abstraction, much like the opposite - communism.
You see - there is a problem with externalities (negative and positive).
But our problem is that when our government starts to regulate in places where regulation is needed (e.g. pollution control) it does not know when to stop.
Even worse: regulation (correctly made or not) in one area forces your hand in another.
In this case - because oil products get huge subsidies in form of tax going for all the military hardware in the Middle East (and note - not from gas taxes), you have to subsidize research into efficient transportation because market does not favor it.
Remember oil spike? Market was alright pushing huge sums of money into alternative energy research.
Russians began it by charging to use "their" resources
Actually, Russians started to charge for the ride to the station which is completely understandable considering costs involved.
Summary is rather inflammatory by calling it "charging for resources".
Yea, the toilet thing is a really petty retaliation. But this is the era of asymmetric warfare after all...
What about Metaram DDR3 modules? Hynix started to roll them out.
Basically they make largish modules using lower capacity (but much cheaper) chips, buffered in such way that to the system they look like slightly slower high capacity ones.
There was a news story few days ago, saying that Intel just certified them.
ICEs are cheap
No, they are not. You forgetting added wight (need a bigger and, alas, heavier battery to accelerate), loss of cargo space, cost of maintenance, cost of gas, even weight of gas. It all adds up in a bad way.
You see, hybrids don't make much financial sense now (go see Toyota dealership's parking lot, after Prius hype is over). And with better battery technology they won't make sense tomorrow. They are simply a temporary kludge on the way to pure electrics.
About "drastic" improvements: current battery technology is already good enough for second, commute only car. With gradual improvements (or BetterPlace like schemes) it will become a primary car.
No, tomorrow they (Volt like hybrids) probably will not make much sense either.
You see - with "higher capacity / cheaper batteries in future cars" you won't need internal combustion engine in them at all.
"native Kindle books" are not tied to the Amazon account.
Kindle's native format is Mobipocket. There are free converters from pdf, html, word. Simply convert, connect Kindle to your computer via usb and upload.
What people seems don't understand is that kindle and kindle like devices are mp3 players for books. And there are not only free classic books out there in the internet. If you know where to look you can find pretty much any book (e.g. pirated Potter books appear about 10 hours after release).
So Kindle is not an "attack on the first sale doctrine" but attack on the whole publishers racket.
And yes, I have Kindle 2. And I love it!
1. Find me a motherboard that has 8 RAM slots that doesn't require expensive ECC and/or Registered memory
You don't need 8 slots. Just use metaram.
3. Find me a computer that can use a RAM drive that can be persistent through reboots without having to save the contents to something else.
That's the thing - you don't need to. Your application should be able to use it naturally. Just make sure you have a modern CPU which can handle it without performance degradation due to paging complications.
That's all in the game. This is the case with any risk measure, alone or in combination with others. The worst-case scenario is the Dies Irae; you cannot take it into account if you want to do any business at all.
Here's the thing: you cannot put it "into the game" when you use incorrect distribution (as they did with VaR). Otherwise pretty much everything (including the world end) can be taken into account given the right price (and possibility of payments in hard currency). The main problem will become the "gambler ruin" (in real "act of God" situations) which will drive premiums even higher than the distribution calls for. Plus, by all accounts, 2008 was not the year the world ended (only seemed that way).
So VaR as used by them was not even a proper "risk measure" - that was one of mistakes they did thinking that it was.
Sure, if I were working for a Wall Street bank and used a more realistic model I would get my ass fired in no time for what it would look like being so risk averse. But that's a systematic problem with businesses today.
The point was (and I am sorry if it didn't come across clearly), that though banks are better diversified on the micro level, this leads to them holding similar portfolios, introducing systemic risk. That is, banks, as an aggregate, are difersified poorly as a result (among others) of Basel-induced VaR contsraints.
Ok. So they diversified they asses to such extent that overall financial system "connectedness" grew high enough to make it unstable.
The rational players are doing just fine.
I smell a common mistake here: "rational players" and "lucky players" are indistinguishable at this point.
That's the problem with markets.
Hell, even with Buffet it is hard to be all one hundred percent sure that he's indeed a genius and not just one really-really lucky dude.
VaR is a pretty decent risk measure on a micro scale.
No, it's not. More precisely it is but only for 99.9% of days. That's Taleb's point. You look at the VaR for 3 years in a row and it's always right, you grow to trust it, and then the big whoop comes.
The real problem with it is that VaR constraints tend to make banks less diversified
Actually it's other way around. It's the property to diversify as much as you please that VaR provides. That's why improper calculation of risks involved caused such a wide spread of problems.
What the financial markets need is a better understanding of the causal links between risks and to price the returns on investments accordingly.
And that's the problem - what are you proposing can't be done. Even fully deterministic systems can display unpredictable behavior (e.g. 3 body problem). Chaotic systems with random events thrown in WILL display that "fat tail" thing even when you know ALL causal relationships in them.
Taleb made his money by not correctly estimating what's the real risk situation is, he made it by betting on his belief that others are off by a long shot.
That's "climate", not "weather".
Weather is chaotic - meaning 2-3 weeks prediction tops. Climate is too, but horizon is much further away.
Hell, solar system itself is chaotic, meaning you cannot predict positions of planets very far into the future. But thousands of years - no problem.
That's so cool!
If I remember correctly, Nikola Tesla invented an apparatus to wreck buildings from within by finding building's resonant frequency and pounding at it.
Kinda like modern mass dampers, but active and, well, not damping much...
Actually, that's not a bad idea for a covert channel.
No, the point of it is that it's a covert channel.
Sure, but write leveling algorithms are essentially black boxes.
The thing is - to avoid trashing SSD when it's almost full you need to actually move data around - i.e. move "stable" data sectors from "fresh" cells into highly used ones. So the obvious problems are:
1. What if my application is such, that it will "chase" the same blocks that the controller targets for moving?
2. When do you do that background moving (reading&writing) thing? If I pound the disk continuously it will suddenly lose 50% write performance (which is not too hot to begin with).
So the numbers for performance and reliability are very application based (true - situation is the same as with mechanical drives but much less predictable). I.e. you don't know how well particular drive model will do until you actually try it (and for a long period of time)
So you select an SLC and sit with your fingers crossed.
Or something like this:
1. You own a battery (and take good care of it).
2. When you go for a trip, first thing you do is stop by the local "U-bat" and install leased battery (your's going into their cool basement with complementary recharge)
3. You drive as far as you please, changing from one leased battery to another.
4. After returning home you stop by aforementioned "U-bat" to install your battery back.
If for some reason you skipped the step 2 the only downside is that there will be additional shipping charge.
So essentially, the only things that are needed: discharge meters on leased batteries and crypto lockup on your battery so that it cannot be used with any other car.