The government (in the USA anyway) has at LEAST the following: Your full name, birth record, race, eye color, hair color, parents names and IDs, your social security number, address, drivers license number, license plate, vehicle VIN number, vehicle registration number, insurance information, bank account numbers, credit account history, mortgage information, phone number (if you have ever included it on a form or called them from home, but they can get it on request anyway if you haven't), tax history, employer name(s), payroll information, fingerprints (from birth, typically elementary school age in most states, and adulthood if you've ever been to a police station or filed them voluntarily), your dental records and medical records (by request of a judge or coroner), military ID and rank (if any), and the list goes on. You act as if the government is a single, monolithic entity. It's not.
"The Government" is a hodgepodge of agencies with mutually contradictory goals and aims, most of whom would sooner throw rocks at each other than cooperate. This is, perversely, a good thing.
Why? Because although "the government" may know a lot about you, it doesn't know all of that in any one place. There's no single database -- yet -- where you can sit down, CSI-style, and bring up any citizen's dossier. Your local police department knows your name, address, and how many parking tickets you've gotten this year, but they don't have access to your tax information from the IRS. (And the IRS is actually pretty snarky about not sharing information casually; if I had a dime for every time one of my LEO buddies bitched about the IRS making them jump through hoops, I'd be a rich man. I guess there's honor among thieves or something.)
This is the way the system is supposed to work. (Well, I'd like to see the size of the bureaucracy cut down dramatically, but that's a different topic.) In order for the bureaucracy to function, it needs to know a certain amount about you. But different agencies need to know different things. As long as the data is kept compartmentalized -- as it is, in large part, today; owing less to design than simply because it's a really hard problem to correlate it all -- it's not a mortal threat to privacy.
It's when you start to get all that information put into a single database, and where there's a natural primary key that allows the database to be easily searched and information to be linked (why do people get paranoid about SSNs? Because they're the obvious choice for a primary key), that you start to get really Orwellian. With minor exceptions, we don't have anything like that in the U.S., although there are a lot of people trying.
Bittorrent (at least current implementations -- not sure about earlier ones) uses TCP only, no UDP.
But you're right, tunneling its connections over UDP would stop the reset problems, and encrypting it would prevent deep-packet inspection. However, the nature of Bittorrent requires creating a lot of simultaneous connections; the overhead of putting each one in its own VPN might be significant. But it's probably the best off-the-shelf solution.
You'd just need to convince all the BT client programs to include it and roll out compatible updates quickly, and hope that the result isn't 5 or 6 mutually-incompatible implementations.
That solution as written doesn't work, and even if it did, might still screw up the connection (because you want to un-set the RST flag, not throw away the whole packet). Also, some people have indicated that Comcast is doing more than just forging RSTs, they are also eating packets along the way, so it's not a silver bullet.
No, they won't, and it's not as easy to filter out RST packets as you seem to think it is. TCP RST packets are handled by the network stack, which is part (usually) of the OS kernel, not userland applications. So to block them, you'd need to change things at fairly low levels of the system, and you'd have to change them at both ends of the connection -- the client requesting the content and also the server supplying it (in the case of web traffic that you're trying to block).
And you really do not want to start telling everyone to configure all of their internet-facing machines to ignore all RST packets. That would create a hell of a mess.
Some ways I've been thinking you could get around the RST-injection MITM attack: set up an encrypted tunnel between the two ends of the connection, and ignore all RST commands that aren't sent inside the tunnel. This would require modifications to the network stack on both ends, though, because you'd have to set it up to specifically ignore RST packets on encrypted connections. (So basically you'd need to have a way of ignoring RST packets on particular connections, but then a way for the higher-level application that did the decryption to reset the connection when it received a correctly-formed request INSIDE the encrypted tunnel.)
Yeah, it works better. Sending a RST packet closes the TCP connection. Just eating the packet would cause the computer to resend it, creating more traffic on the network. The forged-RST attack is "fire and forget." You identify a TCP connection that has bad traffic in it, and then you target the connection. It doesn't require matching every packet, you can instead look for patterns of packets that indicate types of traffic you dislike, and then just terminate it, and move on to the next connection. It may use deep-packet inspection, but it's not a 'packet blocking' attack. It's better, because it avoids having the computers retransmit packets that just contribute to the traffic you need to screen.
It's a fairly insidious way to block traffic, which is why the Chinese do it. Frankly it's a fundamental weakness of TCP: it wasn't really designed to cope with hostile intermediate nodes. (Flaky ones, sure, but not hostile ones.) You could configure your computer to reject RST packets, but then you'd end up leaving connections open all over the place and cause all sorts of other problems. It's not something that you can trivially work around.
One caveat is that the electric transmission capabilities are not up to the task of something like this. Yes, in theory there is sufficient power generation capacity, but moving there isn't a strong enough transmission infrastructure to move this capacity around to where it would be needed. That's one of the reason there's so much extra generation capacity to be found.
Sure there is. The power companies move tons of power into urban areas during the day, and significantly less at night. Also, you can run power-transmission lines at higher capacities during the night than during the day (since it's cooler, you can run more power through high-tension lines before they heat up and sag enough to hit the treetops). There is a lot of excess off-peak grid capacity.
If you moved everyone to time-sensitive, spot-market billing for their power, people would probably move very quickly to only charging their cars during the night. (You'd just want to work with the charger manufacturers to make sure that timers were standard equipment, so that people would have the ability to do it easily.)
Sort of. The "zero" part refers to emissions from the fuel system, which are a significant greenhouse contributor. (My chemistry knowledge here is a little dim but apparently the aromatic hydrocarbons are actually worse when released into the atmosphere than the combustion products, hence all the vapor-recovery nozzles and such.)
The tailpipe emissions have to meet stringent standards as well, SULEV I think, but it's the fuel-system that gets them the "Partial zero" instead of just being 'Super Ultra Low' or whatever.
Yes, but you get a semi-infinite. Unless you divide zero by a partial zero, at which point the result is slightly undefined. I had several professors who were noted for their application of vague number theory.
While I agree with your general points, I think there is a valid role for the government in providing services that cannot (or that we don't want) be limited to only those who have paid for them. In economics terms, those goods for while the 'free-rider problem' is hard to solve.
I think GPS falls into this category. Putting the GPS constellation up was very expensive. Putting it up there, and also building in some capability that made the signal only useful to those who had paid a subscription fee, would have been harder still. (And with a subscription service, I doubt it ever would have become popular enough to pay for itself. I own three GPS units, but I doubt I'd own any if they required a subscription service.) So rather than having no GPS system at all, or a crippled one, you accept that it's something that's useful to society in general and pay for it out of taxes, and allow everyone to use it.
Obviously this is a dangerous game -- it's easy for corrupt politicians to expand the scope of government if not kept in check constantly -- but there are lots of situations where it's the most efficient and effective solution to a problem, to use public funding.
The current scientific journal system is beyond corrupt, and needs to die. However, privatizing all scientific research would be a disaster. First, although you would think that corporations (not having any pesky biological lifespan) would take the long view and invest in basic research, for the most part, they don't. The market favors next-quarter gains, not decades- or centuries-long strategy.
Second, it wouldn't be very healthy to have the majority of our scientific knowledge locked up by corporations who have no interest in it except insofar as it can be monetized or used to gain a competitive advantage. (Hard to put together unified theories when IBM knows half of what's known in a field, Microsoft knows the other half, and they don't speak to each other.) We suffer as a whole, if new discoveries aren't made public. The current academic system (where the currency is basically prestige, rather than cash) encourages dissemination of new discoveries. A more market-driven one would not.
The market economy is a great thing, but there are some areas in which the outcomes it produces may be non-optimal from the point of view of people actually living in the market. Solving the free-rider problem, either when it's not possible to charge for a good, or you don't want to charge for a good, is one of the legitimate functions of a democratic government.
I'm sure if China really wanted to screw with our economy they would "leak" any microsoft code they had- that would put an end to Bill's empire quite efficicntly and all those billions would need to work elsewhere. certainly that would cause the economy to shift a little, wouldnt it? No, it wouldn't. The Windows source code has been leaked before. It caused more problems for OSS projects than it did for Microsoft. Not sure about today, but you used to be able to find it trivially on most P2P networks. Since it's not legitimately public, you can't do much of anything with it -- not if you're a legitimate business/project, anyway. And it provided a good way for Microsoft to question OSS projects that challenged their monopoly (like ReactOS); if anything played too nicely with actual MS software, they could accuse it of having leaked MS code inside, and force an expensive code audit (at best), or sue them into oblivion (at worst).
Anybody on the black hat side (or in countries that don't care about IP) who wants the Windows source already has it. The Vista source will eventually leak, too. Microsoft doesn't really care, because it's pretty hard for anyone to actually use the leaked code to compete with them in their core markets. No competitor would want to touch it, because it would open them up to lawsuits and death-by-code-audits. No OSS project wants it, for the same reasons. People don't even want to work with developers who've seen leaked MS code; the stuff's like poison. And users don't care about access to a hunk of source code -- all they want is binaries, and anyone who wants to can just get a pirated copy of Windows.
So the Chinese could do that, but it wouldn't do anything to harm Microsoft. The Windows code is out there for the taking already.
It costs Google nothing to link to another site, while it costs them to syndicate wire stories. OTOH, if Google displays the story they can keep the ad revenue. Since there are lots of sites around that do nothing but republish wire service reports in order to bring in ad revenue, it stands to reason that the ad revenue is more than enough to pay for the wire service.
Therefore, what Google is setting up is a paying proposition. They're end-running all the eyeball-grabbers.
Rereading the GP, I think he may have been trying to make the same point I was going for; I didn't realize that Microsoft's revealing of the Windows source code to governments was part of the "Shared Source" program (I thought "Shared Source" was just the code name for their quasi-OSS stuff, available to anyone).
They don't need Microsoft's shared source program -- Microsoft provides the Windows source code to foreign governments pretty freely (assumedly under NDA, but it's not like that's really going to mean much to the PRC government). I'm sure the Chinese already have it. All they'd need to do is threaten do dump Windows on all their bureaucrats' machines if MS didn't pony up and let them comb through the source. They hand the source over to their own security people for verification, to make sure it's not backdoored... but if those same security people find a fault, there's no reason why they'd notify Microsoft (or the US government) about it.
The way Microsoft provides access to source code favors whoever is (1) big enough to twist MS into give them the source in the first place, and (2) with the most resources to assign to combing through it for vulnerabilities. Just for manpower reasons alone, I suspect China wins #2 handily.
And the smarter nations have been saying that OOXML features should be harmonized into ODF. In all seriousness, what "features" does OOXML offer over ODF, besides the poison-pills introduced by Microsoft ("do margins like Word 97" and similar)? It seems like they're both doing the exact same things.
He wanted to do it on his own terms in order to save face. The americans nuking of japan was an atrocity and cannot be defended. The Japanese Emperor's sacrifice of hundreds of thousands of his own people simply because he was too obsessed with "saving face" to surrender was an atrocity and cannot be defended.
Cute. So the other 99.99% (don't question the number) of 8th graders who don't have an 'accessory gmail account' are worthless/sub par? We're in for a grim future then:) That's nothing new, really. I think 99.99 percent is probably a stretch, but out of any 100 randomly selected children, I strongly suspect that 90 will live their entire lives completely at the whims of various higher authorities, never bothering to seriously challenge or question them. That's not really a commentary on our society in particular as much as it is human nature in general; I suspect you could go back 2000 years and see basically the same things.
If on your way out of a store the proprietor asks to look in your bag and you think that is an invasion your remedy is to return the items you bought, get your money back and leave.
This is ridiculous. If someone says "do x if you want to remain on my property," and you don't want to do x, the remedy is to leave the property. If you're walking out of a store, carrying nothing but your own goods (which you just bought, so they're now yours), I don't see how the store owner ought to be able to force you to go back into the store and return your purchases, before leaving. The purchase transaction is complete when you walk away from the register (well, technically it's probably complete somewhere around the point where you hand them money and they give you change back), it's not conditional on you doing anything afterwards. That'd just be stupid; they've transfered the goods to the purchaser in exchange for money -- the shop owner can't decide unilaterally to renege on that and force the customer to reverse the transaction.
Furthermore, in most states, before you are guilty of trespassing, you have to be asked to leave (or know you shouldn't be there in the first place) and then not do so. If you don't let them leave, they can't be trespassing, because they couldn't leave. There's no criminal intent there -- they could be acting completely in good faith, but they're being denied the opportunity to not commit a crime.
My wife is a special officer and two of my good friends are full-timers. They would kick the arse of any shop staff that tried to hold a customer without having observed scope/scone. Unfortunately, in the U.S., it's quite common for stores to force you to show a receipt before they'll let you leave. Wal-Mart, most "wholesale clubs," many big-box stores, and an increasing number of electronics retailers do it. I guess they've found it's cheaper to hire some goon at minimum wage to harass people on their way out the door -- after they've made their purchases! -- than to implement a theft-prevention system or to hire enough people to actually track shoplifters and target them specifically.
Well said, and I hope I remember your post in the future, but did you really compare Chomsky and Coulter in the same sentence? Yeah, I knew that was a bad comparison when I wrote it, but nobody came to mind as being a recognizable "Chomsky of the right," at least that I knew of, who would fall into the same category of 'conservativism' as Fox News (authoritarian-conservativism, IMO). Most of the people I think of who are in the same intellectual realm as Chomsky and arguably right-wing fall into different 'flavors' of what's broadly termed conservativism in the U.S. (Meaning, if you didn't collapse the political realm into one dimension, they're not very close at all.)
Historically, you could point towards someone like Joseph de Maistre as an 'intellectual authoritarian conservative,' but it's a pretty sparse field these days... the "godfather" of modern neoconservativism is usually said to be Irving Kristol, who is well-spoken and reasonably well-known (although not like Chomsky), but I don't think that Fox's brand of conservative ideology and Kristol's neoconservativism are exactly the same thing. But I guess he probably would have been a better choice to name-check than a blowhard like Coulter.
A submarine does not need a carrier battle group. The point of a sub, is a stealthy platform for launching missiles or for sneaking up on other vessels undetected. And for killing other submarines.
A sub doesn't need a carrier battle group, but naval strategists (in the US, anyway) have decided that carrier battle groups need submarines -- not for offensive capability, but as the best defense against an enemy submarine. So they pair each battle group with an attack sub in order to mitigate the risk that an enemy could sneak up, pop off one torpedo, and destroy the whole thing.
in case the linux-side DHCP server didn't understand the packets, couldn't it send some kind of error
No really. It could maybe send a NAK but that wouldn't help. The best thing it could do is ignore the unicast option, reply with a normal broadcast packet and hope that the packet gets there. Unless the folks inexplicably designed the network with a broadcast-only dhcp server where the broadcasts couldn't reach the requesting host, the response would most likely get there. From my understanding, where you hit problems is if you have a network that does DHCP over multicast, not broadcast, and where broadcast packets aren't propagated through your routers. If this is the case, the DHCP server will either respond to the client's request with a broadcast packet, which the routers will drop, or it will respond with a multicast packet that will reach the client, but the client may not accept (which is what Vista apparently does).
There are lots of reasons why you'd want to do DHCP over multicast instead of broadcast -- it lets you have one centralized DHCP server, and not have to implement it at every downstream router (but would still let you use the routers to prevent broadcast traffic from growing out of control).
Bush touts the media as having a liberal bias. In reality the media is strongly conservative, Fox news, rather than being the conservative voice, is just outright fascist, and most people still believe everything they hear on their particular brand of news. Expect very little protest as this most recent step in the massive defecation on our rights probably won't make a sub note in the evening news. The fact that Fox News is so popular is indicative that the mainstream (non-Fox) media is, in fact, not as conservative as the majority of viewers want(ed) it to be. Since I think it's a fair assumption that most viewers want media that shares their own biases, we can say then, that the non-Fox MSM is more liberal than most viewers.
On some hypothetical "absolute scale" of liberal/conservative, it might be true that CNN or ABC is 'conservative' and Fox only more so, but in reality there is no absolute scale. Everything is relative to something else: either the citizenry at large, or to the consumers who affect a particular market.
To Noam Chomsky, it's probably true that CNN is very conservative. On his own scale, he's the zero point, and CNN is right of him, and Fox even further right of that. To Ann Coulter, they probably both read as rather leftist, because she's her own zero point and they're both left of her. Depending on which opinion poll you want to believe, the "American public" is somewhere else on the spectrum, and various news sources are 'conservative' or 'liberal' relative to that.
The only borderline-objective source for normalcy seems, to me, to be what the market actually produces in response to consumer desires. It's easy to lie on an opinion poll to make yourself look or feel good, when you're not spending your own money or time. But the market is a good measure of what people actually do; and people abandoned CNN in the late 90s and early 2000s to watch Fox News instead. That's an indicator to me, that the public is actually quite -- perhaps frighteningly -- conservative.
So your assertion is that it is hypocritical for anyone with access to money or power to ever make a statement supporting charity or paying one's taxes? No it's only hypocritical for someone who doesn't pay taxes -- or runs an organization that doesn't pay taxes -- to make statements about others who also don't. It only becomes more comical when you consider that the Vatican itself is basically a tax haven, but for a single organization.
Humm, come to think of this, I think the Pope's own book has some advice for situations like this. I think it goes something like "He that is without sin...first cast a stone".
Then again, Japan does not have military bases in a bunch of Middle Eastern countries, or give billions in aid to a certain Mediterranean country... I think it's less that, than they have a public that's less reactionary than in the U.S. It's not as if Japan is exactly a stranger to terrorism -- both of the international and home-grown variety. They just didn't go absolutely batshit insane as a result of it.
Slashdot Mirror
"The Government" is a hodgepodge of agencies with mutually contradictory goals and aims, most of whom would sooner throw rocks at each other than cooperate. This is, perversely, a good thing.
Why? Because although "the government" may know a lot about you, it doesn't know all of that in any one place. There's no single database -- yet -- where you can sit down, CSI-style, and bring up any citizen's dossier. Your local police department knows your name, address, and how many parking tickets you've gotten this year, but they don't have access to your tax information from the IRS. (And the IRS is actually pretty snarky about not sharing information casually; if I had a dime for every time one of my LEO buddies bitched about the IRS making them jump through hoops, I'd be a rich man. I guess there's honor among thieves or something.)
This is the way the system is supposed to work. (Well, I'd like to see the size of the bureaucracy cut down dramatically, but that's a different topic.) In order for the bureaucracy to function, it needs to know a certain amount about you. But different agencies need to know different things. As long as the data is kept compartmentalized -- as it is, in large part, today; owing less to design than simply because it's a really hard problem to correlate it all -- it's not a mortal threat to privacy.
It's when you start to get all that information put into a single database, and where there's a natural primary key that allows the database to be easily searched and information to be linked (why do people get paranoid about SSNs? Because they're the obvious choice for a primary key), that you start to get really Orwellian. With minor exceptions, we don't have anything like that in the U.S., although there are a lot of people trying.
Bittorrent (at least current implementations -- not sure about earlier ones) uses TCP only, no UDP.
But you're right, tunneling its connections over UDP would stop the reset problems, and encrypting it would prevent deep-packet inspection. However, the nature of Bittorrent requires creating a lot of simultaneous connections; the overhead of putting each one in its own VPN might be significant. But it's probably the best off-the-shelf solution.
You'd just need to convince all the BT client programs to include it and roll out compatible updates quickly, and hope that the result isn't 5 or 6 mutually-incompatible implementations.
That solution as written doesn't work, and even if it did, might still screw up the connection (because you want to un-set the RST flag, not throw away the whole packet). Also, some people have indicated that Comcast is doing more than just forging RSTs, they are also eating packets along the way, so it's not a silver bullet.
No, they won't, and it's not as easy to filter out RST packets as you seem to think it is. TCP RST packets are handled by the network stack, which is part (usually) of the OS kernel, not userland applications. So to block them, you'd need to change things at fairly low levels of the system, and you'd have to change them at both ends of the connection -- the client requesting the content and also the server supplying it (in the case of web traffic that you're trying to block).
And you really do not want to start telling everyone to configure all of their internet-facing machines to ignore all RST packets. That would create a hell of a mess.
Some ways I've been thinking you could get around the RST-injection MITM attack: set up an encrypted tunnel between the two ends of the connection, and ignore all RST commands that aren't sent inside the tunnel. This would require modifications to the network stack on both ends, though, because you'd have to set it up to specifically ignore RST packets on encrypted connections. (So basically you'd need to have a way of ignoring RST packets on particular connections, but then a way for the higher-level application that did the decryption to reset the connection when it received a correctly-formed request INSIDE the encrypted tunnel.)
Yeah, it works better. Sending a RST packet closes the TCP connection. Just eating the packet would cause the computer to resend it, creating more traffic on the network. The forged-RST attack is "fire and forget." You identify a TCP connection that has bad traffic in it, and then you target the connection. It doesn't require matching every packet, you can instead look for patterns of packets that indicate types of traffic you dislike, and then just terminate it, and move on to the next connection. It may use deep-packet inspection, but it's not a 'packet blocking' attack. It's better, because it avoids having the computers retransmit packets that just contribute to the traffic you need to screen.
It's a fairly insidious way to block traffic, which is why the Chinese do it. Frankly it's a fundamental weakness of TCP: it wasn't really designed to cope with hostile intermediate nodes. (Flaky ones, sure, but not hostile ones.) You could configure your computer to reject RST packets, but then you'd end up leaving connections open all over the place and cause all sorts of other problems. It's not something that you can trivially work around.
One caveat is that the electric transmission capabilities are not up to the task of something like this. Yes, in theory there is sufficient power generation capacity, but moving there isn't a strong enough transmission infrastructure to move this capacity around to where it would be needed. That's one of the reason there's so much extra generation capacity to be found.
Sure there is. The power companies move tons of power into urban areas during the day, and significantly less at night. Also, you can run power-transmission lines at higher capacities during the night than during the day (since it's cooler, you can run more power through high-tension lines before they heat up and sag enough to hit the treetops). There is a lot of excess off-peak grid capacity.If you moved everyone to time-sensitive, spot-market billing for their power, people would probably move very quickly to only charging their cars during the night. (You'd just want to work with the charger manufacturers to make sure that timers were standard equipment, so that people would have the ability to do it easily.)
Sort of. The "zero" part refers to emissions from the fuel system, which are a significant greenhouse contributor. (My chemistry knowledge here is a little dim but apparently the aromatic hydrocarbons are actually worse when released into the atmosphere than the combustion products, hence all the vapor-recovery nozzles and such.)
The tailpipe emissions have to meet stringent standards as well, SULEV I think, but it's the fuel-system that gets them the "Partial zero" instead of just being 'Super Ultra Low' or whatever.
While I agree with your general points, I think there is a valid role for the government in providing services that cannot (or that we don't want) be limited to only those who have paid for them. In economics terms, those goods for while the 'free-rider problem' is hard to solve.
I think GPS falls into this category. Putting the GPS constellation up was very expensive. Putting it up there, and also building in some capability that made the signal only useful to those who had paid a subscription fee, would have been harder still. (And with a subscription service, I doubt it ever would have become popular enough to pay for itself. I own three GPS units, but I doubt I'd own any if they required a subscription service.) So rather than having no GPS system at all, or a crippled one, you accept that it's something that's useful to society in general and pay for it out of taxes, and allow everyone to use it.
Obviously this is a dangerous game -- it's easy for corrupt politicians to expand the scope of government if not kept in check constantly -- but there are lots of situations where it's the most efficient and effective solution to a problem, to use public funding.
The current scientific journal system is beyond corrupt, and needs to die. However, privatizing all scientific research would be a disaster. First, although you would think that corporations (not having any pesky biological lifespan) would take the long view and invest in basic research, for the most part, they don't. The market favors next-quarter gains, not decades- or centuries-long strategy.
Second, it wouldn't be very healthy to have the majority of our scientific knowledge locked up by corporations who have no interest in it except insofar as it can be monetized or used to gain a competitive advantage. (Hard to put together unified theories when IBM knows half of what's known in a field, Microsoft knows the other half, and they don't speak to each other.) We suffer as a whole, if new discoveries aren't made public. The current academic system (where the currency is basically prestige, rather than cash) encourages dissemination of new discoveries. A more market-driven one would not.
The market economy is a great thing, but there are some areas in which the outcomes it produces may be non-optimal from the point of view of people actually living in the market. Solving the free-rider problem, either when it's not possible to charge for a good, or you don't want to charge for a good, is one of the legitimate functions of a democratic government.
Anybody on the black hat side (or in countries that don't care about IP) who wants the Windows source already has it. The Vista source will eventually leak, too. Microsoft doesn't really care, because it's pretty hard for anyone to actually use the leaked code to compete with them in their core markets. No competitor would want to touch it, because it would open them up to lawsuits and death-by-code-audits. No OSS project wants it, for the same reasons. People don't even want to work with developers who've seen leaked MS code; the stuff's like poison. And users don't care about access to a hunk of source code -- all they want is binaries, and anyone who wants to can just get a pirated copy of Windows.
So the Chinese could do that, but it wouldn't do anything to harm Microsoft. The Windows code is out there for the taking already.
Therefore, what Google is setting up is a paying proposition. They're end-running all the eyeball-grabbers.
Rereading the GP, I think he may have been trying to make the same point I was going for; I didn't realize that Microsoft's revealing of the Windows source code to governments was part of the "Shared Source" program (I thought "Shared Source" was just the code name for their quasi-OSS stuff, available to anyone).
They don't need Microsoft's shared source program -- Microsoft provides the Windows source code to foreign governments pretty freely (assumedly under NDA, but it's not like that's really going to mean much to the PRC government). I'm sure the Chinese already have it. All they'd need to do is threaten do dump Windows on all their bureaucrats' machines if MS didn't pony up and let them comb through the source. They hand the source over to their own security people for verification, to make sure it's not backdoored ... but if those same security people find a fault, there's no reason why they'd notify Microsoft (or the US government) about it.
The way Microsoft provides access to source code favors whoever is (1) big enough to twist MS into give them the source in the first place, and (2) with the most resources to assign to combing through it for vulnerabilities. Just for manpower reasons alone, I suspect China wins #2 handily.
If on your way out of a store the proprietor asks to look in your bag and you think that is an invasion your remedy is to return the items you bought, get your money back and leave.
This is ridiculous. If someone says "do x if you want to remain on my property," and you don't want to do x, the remedy is to leave the property. If you're walking out of a store, carrying nothing but your own goods (which you just bought, so they're now yours), I don't see how the store owner ought to be able to force you to go back into the store and return your purchases, before leaving. The purchase transaction is complete when you walk away from the register (well, technically it's probably complete somewhere around the point where you hand them money and they give you change back), it's not conditional on you doing anything afterwards. That'd just be stupid; they've transfered the goods to the purchaser in exchange for money -- the shop owner can't decide unilaterally to renege on that and force the customer to reverse the transaction.
Furthermore, in most states, before you are guilty of trespassing, you have to be asked to leave (or know you shouldn't be there in the first place) and then not do so. If you don't let them leave, they can't be trespassing, because they couldn't leave. There's no criminal intent there -- they could be acting completely in good faith, but they're being denied the opportunity to not commit a crime.
Historically, you could point towards someone like Joseph de Maistre as an 'intellectual authoritarian conservative,' but it's a pretty sparse field these days... the "godfather" of modern neoconservativism is usually said to be Irving Kristol, who is well-spoken and reasonably well-known (although not like Chomsky), but I don't think that Fox's brand of conservative ideology and Kristol's neoconservativism are exactly the same thing. But I guess he probably would have been a better choice to name-check than a blowhard like Coulter.
A sub doesn't need a carrier battle group, but naval strategists (in the US, anyway) have decided that carrier battle groups need submarines -- not for offensive capability, but as the best defense against an enemy submarine. So they pair each battle group with an attack sub in order to mitigate the risk that an enemy could sneak up, pop off one torpedo, and destroy the whole thing.
No really. It could maybe send a NAK but that wouldn't help. The best thing it could do is ignore the unicast option, reply with a normal broadcast packet and hope that the packet gets there. Unless the folks inexplicably designed the network with a broadcast-only dhcp server where the broadcasts couldn't reach the requesting host, the response would most likely get there. From my understanding, where you hit problems is if you have a network that does DHCP over multicast, not broadcast, and where broadcast packets aren't propagated through your routers. If this is the case, the DHCP server will either respond to the client's request with a broadcast packet, which the routers will drop, or it will respond with a multicast packet that will reach the client, but the client may not accept (which is what Vista apparently does).
There are lots of reasons why you'd want to do DHCP over multicast instead of broadcast -- it lets you have one centralized DHCP server, and not have to implement it at every downstream router (but would still let you use the routers to prevent broadcast traffic from growing out of control).
On some hypothetical "absolute scale" of liberal/conservative, it might be true that CNN or ABC is 'conservative' and Fox only more so, but in reality there is no absolute scale. Everything is relative to something else: either the citizenry at large, or to the consumers who affect a particular market.
To Noam Chomsky, it's probably true that CNN is very conservative. On his own scale, he's the zero point, and CNN is right of him, and Fox even further right of that. To Ann Coulter, they probably both read as rather leftist, because she's her own zero point and they're both left of her. Depending on which opinion poll you want to believe, the "American public" is somewhere else on the spectrum, and various news sources are 'conservative' or 'liberal' relative to that.
The only borderline-objective source for normalcy seems, to me, to be what the market actually produces in response to consumer desires. It's easy to lie on an opinion poll to make yourself look or feel good, when you're not spending your own money or time. But the market is a good measure of what people actually do; and people abandoned CNN in the late 90s and early 2000s to watch Fox News instead. That's an indicator to me, that the public is actually quite -- perhaps frighteningly -- conservative.
Humm, come to think of this, I think the Pope's own book has some advice for situations like this. I think it goes something like "He that is without sin...first cast a stone".