Or even the past European imperial powers like Spain or England.
I happen to be from a past European imperial power (Portugal). It so happens that every European imperial power, except for Spain, decolonized their empires in a pretty constructive fashion. You have to understand that losing a colony means losing a lot of power, so secession wars are inevitable. However, credit is due to the huge diplomatic effort and excellent results that the Brits produced when creating the Commonwealth, or to the excellent relations French kept with their African colonies or to the great relationship between Portugal and Brazil and East Timor, Angola and Mozambique.
For a very fond example of good colonization, Portugal is a prime example. Portuguese colonies were established by mingling with natives, producing unique cultures that flourished on their own with unique features. Look at the average Brazilian, and you'll find a mix of native South Americans with European colonizers. I digress. My generic point is: While the United States haven't, historically, committed large blunders, they are doing their first one right now, with the invasion of Iraq and the political destabilization of a whole region of the globe. Worse, they don't seem to recognize it. Worse yet, and really scaring, is that this is coming from a country that is a developed democracy as they come. It shows how fragile freedom is, and how easily a good people can commit atrocious acts.
And I can't stop being amazed on how self-righteous Americans are, with the ingrained idea that they are, and that they should be, policemen (and jury and executioner) of the World. Your effect outside of your borders is on par with China, veering for the worst. The historical evaluation of your current actions will be much much worse than the historical evaluation of the acts of European colonizers.
That phrase was there with the sole purpose of directing angry replies to it. The fact that you did not dispute any affirmation about the US but instead tried to paint China black says a lot about the current state of the democracy in the States.
Of course China is no clean regime, and that was not my point.
The worst part about the US going into Iraq is the entire world should have gone.
No it shouldn't. Invading countries to 'correct' their regimes assumes you know, from the outside, what is best for them. The result is in plainsight. Iraqis are worse today than under Saddam.
Countries should find their own course. If under revolution then, and only then, should the international community intervene, in order to protect civilians as much as possible.
Countries need to grow socially to adopt socially evolved forms of civilization. You can't cram democracy into countries without basic democratic infrastructures, like press, and without the basic democratic culture.
You're not that far from China. Your current government isn't despotic just because of historic incident. Bush and its minions would gladly rule with an iron fist if it were remotely possible.
As it is, they do it in more subtle ways. Controlling the justice branch of the government and controlling the media as much as possible -- You get more news about Iraq in European tv than in the US.
And, as facts go, you still have the death penalty, you still torture people, you still fire people for having 'wrong' religious views, and worse than anything else you wage wars unprovoked. At least the Chinese haven't attacked any country in modern times.
Maybe you live in a world where the battery, or some other method of storing excess energy generated now, for use when generation isn't happening, doesn't exist...
Or, maybe, you don't know that batteries are the weakest link in power systems today:
Lead-acid batteries, the best ones we have for stationary systems, range between 50% and 70% efficiency in charging and 65%-80% in the discharge part of the cycle. All through the full cycle, you'll get an average 45% efficiency.
The charge density (Ah/Kg) of lead-acid batteries is so low that you'd need half a ton of batteries per household to guarantee 24h reserve capacity, at full battery health.
Lead-acid batteries contain, you know,..., lead. Lead is rather poisonous, and has the tendency to infiltrate water sources being costly to clean out.
Lead-acid batteries, while they degrade very slowly if fully charged, will sulfate when discharged. A lead-acid battery executing full charge-discharge cycles daily will last a couple of months. Have an old car battery around? About 6 years old will do. Discharge it, leave it be for six hours, then try to recharge. It'll be dead (you'd be switching batteries in a couple of years anyhow).
So, we now have 500kg per household, replaced every six months. Is it possible? We could stretch. Is it feasible? No way... Economic? Only if you're insane.
Fuel cells may be a solution. Current batteries don't cut it.
It's so obvious! Solar/wind/tidal aren't available on-demand. They're available on uncontrollable conditions. Power fluctuations aren't acceptable, so a 100% electrical grid (even a personal power grid, duh!) is not possible.
<hyperbole>
Unless you're into random shutdowns whenever a cloud passes by...
</hyperbole>
No - you misunderstand the difference between Large hydroelectric power and Hydroelectric power. A waterwheel, generator and some moving water is all you need for hydroelectric. so anyone on a river with a sufficient amount of water can use this source.
You are talking mini-hydric. Unfortunately, like their larger siblings, they require a river with specific conditions, namely a limited water flow, placing their possible locations far from the consumption location, as I have pointed out.
depends on the generator - a 2k gasoline based generator is fairly low in inefficiency but a 40kw diesel can be very efficient.
Define very efficient. Regular gasoline generators sport around 10% efficiency, diesel generators move to about 15% efficiency. That's a 50% increase, but still on another league when placed against a full-blown power plant which is over 3x more efficient. Note that transportation accounts for a mere 5% losses.
Maybe individual power is really even less feasible than you might realize. The problem is cost to install, not feasibility of the installed system. The addition of sub $1.00 per watt solar coming online in the next 18 months will go a long way to correct that cost to install problem too.
Solar, wind and tidal power, while extremely environment friendly, fail to cope with peaks in consumption. Not to say that they shouldn't be used. They should. It's just that a 100% renewable (minus hydroelectric) power grid is not possible. Maximum advisable ratios for 'random' power sources is 50% installed base. Ideally, it should be complemented with hydric power, which has the ability to use excess capacity by pumping water upstream.
This is all made worse by uncoupled power grids, as you can't compensate randomness in one part of the grid with production from elsewhere.
Maybe someday, with much better efficiency in solar/wind/tidal power and much much better fuel cells or some other energy storage... Not today, though.
Individualized power, which is more feasible than you might realize...
Hmm. How do you answer these concerns:
No hydroelectric power. Large hydroelectric plants are not anywhere close to large consumers, and are too large for any individual consumer, making them unfeasible.
No nuclear plants. Ditto.
Much lower efficiency termic power. Coal/fuel plants improve in efficiency when very large scale. A large fuel plant gets about 50% efficiency, due to heat recovery methods. Individual generators -- the portable kind -- are closer to 10% efficiency.
No load balancing for wind power. Wind power, while efficient, requires load balancing in the grid, like dams pumping water upstream, in order to cope with the fluctuations of power production vs consumption.
No load balancing for solar power. Ditto.
Maybe individual power is really even less feasible than you might realize.
And therein lies your error. Linux wants no one, and doesn't want you. Use it if it fits you, use windows if you prefer. Don't insult other people because of your choice.
That's right. He's a zealot and his commentary on how unsafe Javascript is has nothing to do with backing up his point about turning Javascript off. In fact, his whole post is him just babbling incoherently with no related points whatsoever.
He did say that Javascript is unsafe because it's dynamically typed, whereas Flash is mostly ok. Let's just agree he's not very high in my scale of authority regarding security or computer languages.
What linux is tryng to achieve is light years away from avoiding crackles in audio output. Linux is building a scheduler that maximizes both measured and perceived throughput of a computer. Since there is a fine balance between both concepts, developing an all-around scheduler is prone to a lot of try and error. Schedulers in the 2.6 tree have ranged from acceptable to great, and CFS looks like a great idea. I have yet to try it, though.
No, you're not missing anything. The flow you propose does balance ease of development/debugging with the ability to parallelize query execution. This is, in fact, the architecture I use when pages depend on multiple webservice calls, which usually exhibit a much longer wall-clock execution time than db calls. I just question the amount of parallel database queries that can be executed in most web applications. Most of them are quite simple, and all data can be retrieved in less than three queries, so the gain is not so large.
Anyhow, this is the kind of optimization I'd keep an eye for, but wouldn't introduce in an application without a profiler showing that I need it.
Full async, which was what I thought you were proposing, is still something I will steer clear of if at all possible. Time-dependent bugs are a bitch.
It will make a MASSIVE difference in CPU/RAM usage...
No it won't. It will make zero difference in CPU usage. It will make a difference in RAM usage proportional to the relation between the time to process a request synchronously and the same time asynchronously. Given that, in most applications, this relation approaches 1, the gain is minuscule.
And async is certainly not much more difficult - it's a new way of thinking, sure, something new to learn. But it's not really that difficult!
It's way more difficult to test. You don't know which of your code will be executing when you receive the response from the database, so you can end up with the typical challenges of parallel programming: race conditions, need for locks, time-dependent bugs.
Most databases have async APIs. Postgresql and mysql have them in the C client libraries. Most web development languages, though, do not expose this feature in the language API, and for good reason. Async calls can, in rare cases, be useful for maximizing the throughput of the server. Unfortunately, they're more difficult to program, and much more difficult to test.
High scale web applications have thousands of simultaneous clients, so the server will never run out of stuff to do. Async calls have zero gain in terms of server throughput (requests/s). It may reduce a single request execution time, but the gain does not compensate the added complexity.
Older ones, yeah. They're just pseudo-random generators, with the seed in sync with the car. The car accepts the next 15 numbers in the pseudo-random sequence, and when a valid number is used, it locks/unlocks and re-syncs the pseudo-random generator seed. My car manual comes with instructions on how to manually re-sync car and key if it stops working (for example, a kid clicks the key more than 15 times when the key is away from the car).
Note, however, that having past numbers reveals nothing about the next one. You can have all of my past key identification numbers and gain nothing with it
This attack, however, seems targeted at newer wireless keys. The ones that allow a car to unlock when the owner is near, and start with a dashboard switch without need for physical contact between car and key.
NAT as a technology is irrelevant to where the traffic goes externally, but this discussion is not about where the traffic goes, it's about the validity of the screenshot as a source of any real information, as commented on by the GP.
NAT is irrelevant for the discussion of where the traffic goes. I'm glad we agree on that. It is now a linear deduction that discussing NAT in the context of the packet dump shown in the screenshot is also irrelevant. It is obvious you can deduce where the traffic goes by observing traffic originating or terminating in the target workstation. Regardless of NAT, naturally.
The external interface IP showing up or not is irrelevant. We're not interested in it. Will it show up? Yes if you have no NAT; yes if you are under NAT and observe traffic on the NAT box; No if you are under NAT and observe traffic on the workstation. Do we gain any information of where the traffic is going to by observing the "destination of global incoming packets" presence in dumped packets? Nope. Zero information there for us. Why are we discussing it? Beats me.
My point is that NAT is irrelevant in this discussion. If you're trying to determine whether your PC is sending packets to DoD, Microsoft or whatever, you can disregard NAT.
The 'anomaly' you point out is caused by the fact that the Vista box in the article is the NAT box. It is connected directly to the internet, and outside incoming packets sport the network's external address in the headers.
There you go. Observe that this is a conversation between my PC (10.0.1.76) and slashdot (66.35.150.250). My NAT box (10.0.1.254) is nowhere to be seen in the packet dump. For the objective of determining with whom is my computer trying to speak to, NAT introduces zero noise.
The possibility described in my post explicitly described a scenario of both the Vista box *and* the XP box being behind a NAT box. In that scenario, packets sniffed on the network would *never* contain the global outside destination address of incoming packets. It would be replaced with the local inside destination, being the workstation IP address.
Wrong for the second time. Please go read up on NAT, and stop producing wrong information.
The same logic applies, only you need to observe the source address. The NAT box address never comes up in the IP packet header fields. Only the NATed workstation and the internet server come up. Regardless of NAT, you can always observe the traffic to/from a given machine and identify who is it talking to.
I was guiding myself by the parent post to the one I replied to. Indeed, the article shows packets sniffed from the NAT box. The NAT box, for our interests in this case, behaves as if connected to the net without NAT, turning this issue even more irrelevant.
Not to say that the article isn't ill-conceived and unproven. It's just that it's not because of NAT.
Slashdot Mirror
Vietnam isn't in Africa, last I checked, sor is Israel in the Commonwealth. Read before replying.
I happen to be from a past European imperial power (Portugal). It so happens that every European imperial power, except for Spain, decolonized their empires in a pretty constructive fashion. You have to understand that losing a colony means losing a lot of power, so secession wars are inevitable. However, credit is due to the huge diplomatic effort and excellent results that the Brits produced when creating the Commonwealth, or to the excellent relations French kept with their African colonies or to the great relationship between Portugal and Brazil and East Timor, Angola and Mozambique.
For a very fond example of good colonization, Portugal is a prime example. Portuguese colonies were established by mingling with natives, producing unique cultures that flourished on their own with unique features. Look at the average Brazilian, and you'll find a mix of native South Americans with European colonizers. I digress. My generic point is: While the United States haven't, historically, committed large blunders, they are doing their first one right now, with the invasion of Iraq and the political destabilization of a whole region of the globe. Worse, they don't seem to recognize it. Worse yet, and really scaring, is that this is coming from a country that is a developed democracy as they come. It shows how fragile freedom is, and how easily a good people can commit atrocious acts.
And I can't stop being amazed on how self-righteous Americans are, with the ingrained idea that they are, and that they should be, policemen (and jury and executioner) of the World. Your effect outside of your borders is on par with China, veering for the worst. The historical evaluation of your current actions will be much much worse than the historical evaluation of the acts of European colonizers.
That phrase was there with the sole purpose of directing angry replies to it. The fact that you did not dispute any affirmation about the US but instead tried to paint China black says a lot about the current state of the democracy in the States.
Of course China is no clean regime, and that was not my point.
No it shouldn't. Invading countries to 'correct' their regimes assumes you know, from the outside, what is best for them. The result is in plainsight. Iraqis are worse today than under Saddam.
Countries should find their own course. If under revolution then, and only then, should the international community intervene, in order to protect civilians as much as possible.
Countries need to grow socially to adopt socially evolved forms of civilization. You can't cram democracy into countries without basic democratic infrastructures, like press, and without the basic democratic culture.
Pot, kettle, black.
You're not that far from China. Your current government isn't despotic just because of historic incident. Bush and its minions would gladly rule with an iron fist if it were remotely possible.
As it is, they do it in more subtle ways. Controlling the justice branch of the government and controlling the media as much as possible -- You get more news about Iraq in European tv than in the US.
And, as facts go, you still have the death penalty, you still torture people, you still fire people for having 'wrong' religious views, and worse than anything else you wage wars unprovoked. At least the Chinese haven't attacked any country in modern times.
So, we now have 500kg per household, replaced every six months. Is it possible? We could stretch. Is it feasible? No way... Economic? Only if you're insane.
Fuel cells may be a solution. Current batteries don't cut it.
It's so obvious! Solar/wind/tidal aren't available on-demand. They're available on uncontrollable conditions. Power fluctuations aren't acceptable, so a 100% electrical grid (even a personal power grid, duh!) is not possible.
<hyperbole> Unless you're into random shutdowns whenever a cloud passes by... </hyperbole>
Solar, wind and tidal power, while extremely environment friendly, fail to cope with peaks in consumption. Not to say that they shouldn't be used. They should. It's just that a 100% renewable (minus hydroelectric) power grid is not possible. Maximum advisable ratios for 'random' power sources is 50% installed base. Ideally, it should be complemented with hydric power, which has the ability to use excess capacity by pumping water upstream.
This is all made worse by uncoupled power grids, as you can't compensate randomness in one part of the grid with production from elsewhere.
Maybe someday, with much better efficiency in solar/wind/tidal power and much much better fuel cells or some other energy storage... Not today, though.
- No hydroelectric power. Large hydroelectric plants are not anywhere close to large consumers, and are too large for any individual consumer, making them unfeasible.
- No nuclear plants. Ditto.
- Much lower efficiency termic power. Coal/fuel plants improve in efficiency when very large scale. A large fuel plant gets about 50% efficiency, due to heat recovery methods. Individual generators -- the portable kind -- are closer to 10% efficiency.
- No load balancing for wind power. Wind power, while efficient, requires load balancing in the grid, like dams pumping water upstream, in order to cope with the fluctuations of power production vs consumption.
- No load balancing for solar power. Ditto.
Maybe individual power is really even less feasible than you might realize.$a;<ESC>:x
and go out for coffee while the editor executes.
And now, step off my lawn, you pico loving youngsters!
Oh, so HTML 5 redoes the OBJECT tag. Cool! Progress!
/irony
What linux is tryng to achieve is light years away from avoiding crackles in audio output. Linux is building a scheduler that maximizes both measured and perceived throughput of a computer. Since there is a fine balance between both concepts, developing an all-around scheduler is prone to a lot of try and error. Schedulers in the 2.6 tree have ranged from acceptable to great, and CFS looks like a great idea. I have yet to try it, though.
No, you're not missing anything. The flow you propose does balance ease of development/debugging with the ability to parallelize query execution. This is, in fact, the architecture I use when pages depend on multiple webservice calls, which usually exhibit a much longer wall-clock execution time than db calls. I just question the amount of parallel database queries that can be executed in most web applications. Most of them are quite simple, and all data can be retrieved in less than three queries, so the gain is not so large.
Anyhow, this is the kind of optimization I'd keep an eye for, but wouldn't introduce in an application without a profiler showing that I need it.
Full async, which was what I thought you were proposing, is still something I will steer clear of if at all possible. Time-dependent bugs are a bitch.
Most databases have async APIs. Postgresql and mysql have them in the C client libraries. Most web development languages, though, do not expose this feature in the language API, and for good reason. Async calls can, in rare cases, be useful for maximizing the throughput of the server. Unfortunately, they're more difficult to program, and much more difficult to test.
High scale web applications have thousands of simultaneous clients, so the server will never run out of stuff to do. Async calls have zero gain in terms of server throughput (requests/s). It may reduce a single request execution time, but the gain does not compensate the added complexity.
Older ones, yeah. They're just pseudo-random generators, with the seed in sync with the car. The car accepts the next 15 numbers in the pseudo-random sequence, and when a valid number is used, it locks/unlocks and re-syncs the pseudo-random generator seed. My car manual comes with instructions on how to manually re-sync car and key if it stops working (for example, a kid clicks the key more than 15 times when the key is away from the car).
Note, however, that having past numbers reveals nothing about the next one. You can have all of my past key identification numbers and gain nothing with it
This attack, however, seems targeted at newer wireless keys. The ones that allow a car to unlock when the owner is near, and start with a dashboard switch without need for physical contact between car and key.
Your past posting history on /. reveals an alternative view. You are quite the adept lecturer.
NAT is irrelevant for the discussion of where the traffic goes. I'm glad we agree on that. It is now a linear deduction that discussing NAT in the context of the packet dump shown in the screenshot is also irrelevant. It is obvious you can deduce where the traffic goes by observing traffic originating or terminating in the target workstation. Regardless of NAT, naturally.
The external interface IP showing up or not is irrelevant. We're not interested in it. Will it show up? Yes if you have no NAT; yes if you are under NAT and observe traffic on the NAT box; No if you are under NAT and observe traffic on the workstation. Do we gain any information of where the traffic is going to by observing the "destination of global incoming packets" presence in dumped packets? Nope. Zero information there for us. Why are we discussing it? Beats me.
The 'anomaly' you point out is caused by the fact that the Vista box in the article is the NAT box. It is connected directly to the internet, and outside incoming packets sport the network's external address in the headers.
There you go. Observe that this is a conversation between my PC (10.0.1.76) and slashdot (66.35.150.250). My NAT box (10.0.1.254) is nowhere to be seen in the packet dump. For the objective of determining with whom is my computer trying to speak to, NAT introduces zero noise.
Wrong for the second time. Please go read up on NAT, and stop producing wrong information.
The same logic applies, only you need to observe the source address. The NAT box address never comes up in the IP packet header fields. Only the NATed workstation and the internet server come up. Regardless of NAT, you can always observe the traffic to/from a given machine and identify who is it talking to.
I was guiding myself by the parent post to the one I replied to. Indeed, the article shows packets sniffed from the NAT box. The NAT box, for our interests in this case, behaves as if connected to the net without NAT, turning this issue even more irrelevant.
Not to say that the article isn't ill-conceived and unproven. It's just that it's not because of NAT.