They could try replacing all the keys for all devices as a stopgap, but that's pretty problematic and could well just lead to the same leak happening again.
It wouldn't help much in any case, since HDCP is fundamentally flawed. Extracting 40 particular public-private keypairs (perhaps by electron microscopy or what have you) is sufficient to make a synthetic master key.
We can ignore them altogether. Pirate, build heavily encrypted peer-to-peer networks, migrate from the "official" internet to the "dark" internet, until all that's left on the official internet is corporate advertising. Short of a police state, they can't force us to use the official internet, and the more they tighten their grip, the more people find it worthwhile to slip through the fingers into the unofficial internet.
But that's also rather strange. For quantum crypto to make sense, there must be an adversary who can crack Diffie-Hellman (or the key exchange of your choice), but who isn't able to just physically get on the line and insert a man-in-the-middle device. Even if you have perfect quantum crypto, unless you and your intended other party shares a secret, it's impossible to determine if the key negotiation is between you and your intended other party or with Mallory masquerading as that other party, impersonating you to the other party.
That threat model appears to be irrational. Ordinary inside jobs, industrial espionage, whatnot, can't splice in a man-in-the-middle but they can't break DH either - and if quantum computers become prevalent and DH can be easily broken, one can just shift to some post-quantum PK method like McEliece or NTRU. On the other hand, if the messages are being read by the NSA (who might be able to break DH), quantum crypto won't help, since the NSA could just dig up the cable somewhere and insert a man-in-the-middle device.
That leaves the option where you do have a shared secret. In that case, simply make the shared secret the key for a symmetric key cryptosystem, or if you have the bandwidth, use a one-time pad. QC's only advantage in that case would seem to be that the adversary has no way of stealing the "key" or "pad" without tampering with the cable or compromising one of the endpoints.
(I assume that the communication is being done on a dedicated line, since one can't do quantum crypto over the internet. I also assume that "key exchange" is of the kind where a passive listener has to solve a hard problem to get the key - DH is such a method, but there are others.)
You can't say what would happen in a plurality election based on votes cast in an IRV election. In a plurality with 3 or more candidates, people vote tactically. You can't assume that their single vote would be the same as their first choice in an IRV election.
Alright, but that has no bearing as to the success of AV. If we disregard Plurality, the fact remains that Montroll is both the one-on-one winner and the one with the most first + second place ballots combined. My point was to show that AV is pretty unique in electing Kiss here. Of course you could say that that's not an argument, either, but in the worst case, we can disregard Plurality as it doesn't factor into the IRV/AV vs other methods comparison.
And what does "the worst candidate" mean anyway?
He loses to the other two. If it had been a top two runoff, he would lose no matter who the other candidate had been.
Again, that's crooked thinking. Imagine A is a left wing candidate, but the right has a split vote with candidates B and C both vying for the right. Say that the district is a right wing one. IT may well be that candidate A can individually beat both B and C. But the ideal vote is clearly not to return the left wing candidate for the right wing district. So again you can't just take the IRV votes and munge them into theoretical 2 horse races. It's garbage in, garbage out.
For your first point, consider a left-right spectrum with voters placed along the line and candidates at points close to some of the voters. Then start by "canceling out" the rightmost and the leftmost voter. Do so until there's only one voter left, and the candidate closest to him wins - in other words, the candidate closest to the median voter wins*. If a method satisfies the Condorcet criterion (i.e. picks the candidate that wins all pairups when there is one), then if voters vote honestly (vote candidates closer to them above candidates further out), the candidate closest to the median voter wins. That is called the Black single-peakedness theorem, and implies that a right-wing candidate would win in your example. If you think about it, it makes sense: if the single left-wing candidate A is to beat both B and C, that means either that a majority puts A ahead of both B and C, in which case it's not a right-wing state, or that some voters rank "B above A above C" or "C above A above B", in which case right versus left wing isn't all that matters.
In a more general multidimensional case, the independence of clones criterion is a good yardstick for whether the method is vulnerable to vote-splitting or not. AV passes it, but so do the advanced Condorcet methods like Schulze. Yee diagrams also show that Condorcet methods find the median voter candidate in 2D instances if the voters are distributed in a Gaussian manner around some central point and the distance metric is Euclidean, although that's a more specific case.
For your second point, note that the pairwise margins are significant. If the voters were aiming to bias AV maximally, the margins would be tighter since the final comparison, when there are only two candidates left, is a pairwise one. Moreover, the fact that the AV opposition was able to unify when AV uniquely elected Kiss, but not in the prior election when it picked the same result as the other methods, suggest that AV did pick the wrong winner - that if there was strategy, it backfired. That is no proof, of course, but it does support the margins conclusion. Finally, we may add the first+second vote metric to the heap; if there was strategy, it would be unlikely that the strategy would show a clear winner by both measures, because if each voter crafts his vote to maximally affect the AV outcome, there would be few degrees of freedom left to make someone else the Condorcet winner and Bucklin (first plus second) winner as well.
But you've come up with a non-realistic, very carefully worked out example that appears to do the wrong thing, ever so slightly. But every electoral system is capable of being criticised from some such carefully worked out example. With FPTP you don't even need to carefully work out an example. Remember that FPTP regularly elects people with only 35%ish of the vote.
Unless you live in a place where ballot stuffing is a problem*, AV is indeed better than FPTP. It is wrong, however, to say that AV's failures only occur in contrived examples. AV works as long as the minor parties are so small that they get eliminated before any major parties, but when that is not the case, the parties may be eliminated in the wrong order. Here's an example of that, from the 2009 Burlington, Vermont election for Mayor:
1332 voted Montroll > Kiss > Wright
767 voted Montroll > Wright > Kiss
455 voted Montroll
2043 voted Kiss > Montroll > Wright
371 voted Kiss > Wright > Montroll
568 voted Kiss
1513 voted Wright > Montroll > Kiss
495 voted Wright > Kiss > Montroll
1289 voted Wright
Plurality elects Wright, the worst candidate of the three. IRV elects Kiss even though in a real runoff, Montroll could beat any of the others one-on-one and had more first plus second place preferences than either of the others. The reason why that happened was because other first preference votes obscured Montroll's support and so led to him being eliminated. Because IRV (AV) misjudged Kiss's support and elected him winner, the voters that thought that was a bad idea were able to unite and later replace IRV with ordinary top-two runoff.
* The reason for this is that FPTP can be summed locally: just count first preferences and send the array to the central location. AV has to be counted centrally so the actual ballots have to move further, unless you have the required infrastructure in place to send messages back and forth about whom to eliminate at each stage.
So party A has the most first choice votes, and party A has the most second choice votes. But party B gets in. Instead of making 49% of people completely happy and 48% slightly happy, you're making 48% completely happy and 33% slightly happy. Why are you giving the final say to the second choice of those who have voted for the least popular candidates?
That's because AV is not a very good single-winner method. What you want is something more like what Wikimedia uses - a Condorcet method, where each candidate is counted as beating the candidates ranked below it, and the candidate that beats every other one-on-one (like in sports) wins. Unfortunately, it's too radical (with a very few exceptions, no such method has been used for governmental elections) and so it has absolutely no chance even in situations where using a single-winner method would make sense (like electing a president or a party leader).
For your example, a simple count-the-winning-side Condorcet method would give:
A preferred to B by 49, B preferred to A by 51, B wins and gets 51 points
A preferred to C by 97, C preferred to A by 2, A wins and gets 97 points
A preferred to D by 97, D preferred to A by 1, A wins and gets 97 points
B preferred to C by 69, C preferred to B by 22, B wins and gets 69 points
B preferred to D by 70, D preferred to B by 10, B wins and gets 70 points
C preferred to D by 22, D preferred to C by 10, C wins and gets 22 points
That breaks the fundamental principal of a representative being the voice of his district. If the people living in Town A vote for a Liberal, and he gets replaced by this new AV+ system by a smaller psrty (say libertarian) then he's not really representing that town's views.
True. I would prefer STV with 3-5 member districts, myself, but that wasn't even on the agenda.
Now, you could say that multimember districts like those STV require would weaken the principle of representation because each district would be larger, but that's part of a necessary tradeoff. In order to give minorities more than a zero percent say, one has to weaken the majorities; but at least with STV, the later preferences of those whose candidates didn't win can make a difference. The 50%+/-1 example shows why: in order to give the 50%-1 minority some say, some candidates that would otherwise be allocated to the 50%+1 majority has to go to the 50%-1 minority instead, and there's no way of getting around that. MMP does the tradeoff very abruptly: the constituency part of the parliament is completely majority-based and then the list part patches it up - no majority loss in the constituency part, great loss in the list part. STV does the tradeoff more subtly: a little bit in each district in order to permit PR at all.
If the promise to have a referendum on Alternative Voting is delivered upon, and the electorate are intelligent enough to vote it in, then it will solve this predicament. It will make it always advantageous to vote for the party(s) you prefer, rather than voting tactically for a different party in the hope of keeping the villain of choice out.
AV provides slightly more fair rules, but not enough. To show this most clearly, imagine there are only two parties, and one of the parties gets 50% + 1 of every single constituency. Then half the voters' votes are wasted. A much more fair outcome would give half the seats to the second party, and for that you'll need STV or party list. The Liberal Democrats wanted AV+, which is a combination of AV and party list wherein a party that gets too few constituency seats is awarded top-up seats to compensate. However, AV+ lost in the compromise because the Conservatives don't want proportional representation, and thus they arrived at plain old AV. As Australia shows, it's not enough: Australia uses AV and has a two plus a half party system (Labor on the one hand and National plus Liberal on the other), even with the Senate, which uses proper PR, to counterbalance it.
But if the AV referendum passes, perhaps it will lead to another about true PR. On the other hand, it could also become a reform without reform, discouraging voters and parties from considering better systems because "we already tried that and it didn't help".
The difference is that on the internet, you can use pseudonymity. You can call yourself JSBiff and I can call myself kvezach, and neither of us will know the other person's real name. My different pseudonyms can be limited from each other, as well: if I register as foo on a chess site and bar on a Go site and say the other game sucks on both, nobody will call me on it (if I don't do anything stupid).
In contrast, DRM fails because once you can see it, you can copy it, no matter what pseudonym you (or the creator) may be hiding under. The analogous situation would be that stream-games-from-server thing that has been on Slashdot before: it only shows you what it decides you need to see, so you can't copy the game itself. Well, to stretch the analogy, it would be like a peer-to-peer network version of that, where you don't even know where the game is ultimately stored (i.e. pseudonymity).
The net effect of which would be... letting others decide for you. The same as voting over the line, just in a slightly less predictable fashion. I don't get why it should make such an important difference?
It makes a difference because you can stop at exactly the point you want. To use an American example: optional preference voting would let you rank R, D, and Green candidates without having to care about where you're going to put the Constitution Party. Or a Norwegian one: optional preference voting would let you rank AP relative to FrP without having to care about KSP or NKP. The way Australia's "preference" voting currently works, you would have to rank even the obscure candidates. That's a lot of work, so most people just vote above the line, and so the fine-grained multiwinner system (that has the potential to be proportional not just by party) is reduced to, in essence, a very opaque party list method.
The seventh one is fixed, so we have 38 symbols for 6: 38*37*36*35*34*33 ~= 2*10^9, ends up being 3779 years. That's still too much, but...
There are these pretty crystals inside of the slanted table with all the buttons on it though.
So you have a DHD! If you remember, putting in the wrong address on a DHD just blinks the buttons and nothing happens. Therefore, the activation sequence takes much shorter time than a minute. Let's say that you could get it down to a second with the proper robotics. Then you could exhaust the space in 63 years. Also, the DHD only has 38 symbols, and you can extract about 50 addresses from each if you know how. If you could figure out how, you could make a robot that steals the addresses, tries them all, and repeats for those that actually open.
You have to get to 70% of c just to make a 500-ly journey feel like 500 years, and even closer if you want it to feel quicker. Good luck getting that close to c without antimatter. Good luck getting that close to c even with antimatter.
I can turn it on, but I haven't found a combination of these freakin' buttons to do anything useful.
JW, this is David Lightman. David, this is JW. Well, David, why don' you tell JW about this amazing trick you've discovered on your computer? I think you're calling it "wardialing" or something similar...
I remember a story a few years ago saying congressional approval levels are always dismally low and yet historically about 90% of congressmen are re-elected. Why? Because the majority of people absolutely hate Congress but think their personal representatives are doing a pretty good job.
And perhaps also gerrymandering. When the representatives can pick their voters before the voters vote, then it's no wonder the incumbency rate is so high. Contorted districts like this one aren't drawn just for the amusement of the mapmakers.
It does seem that they were very focused on being able to extract the oil rather than just stopping the leak. Now, I'm not an engineer, but could their desire for continued extraction of oil have delayed their plans, made the stack more complex?
In any case, we'll see whether it works. Hopefully it'll at least buy them enough time to drill relief wells.
really, there's no way around this that can't also be worked around by the spammers. Every single step is met by counter action and evasion. The only thing that works is jail time.
How about the idea of proof-of-work with price discrimination? Unknown domains start at a fairly high level, so it takes a long time to send mail. If the user or domain has sent a few good mails, it's "trusted" (but using a robust trust metric so that spammers can't just trust each other), and then sending mail is fast. If the user gets compromised, he drops on the trust network and suddenly it's expensive to send mail again.
It's kinda a hack because you're explicitly wasting processing (or memory access) power to limit something that's otherwise unlimited, but it could work. That is, it could work once someone waves his wand and establishes that distributed trust network.
Botnets would initially circumvent this, but the zombies would get their trust degraded and so would send spam very slowly, making them unsuitable. Thus the spammers would have to ever chase new hosts. The delay function can be structured so that "stop and go" (send some spam until at worst trust, then let the user get back to good trust, then send more spam) doesn't work.
The closest thing I could find that would fit his description is this, but you have to cut through a lot of noise about how the Bilderbergers/NWO/international Jewish conspiracy is controlling the world to get at what it's actually saying. Basically, it seems to have mixed up a person's forcible hospitalization, her statements against the incorporation of Norway into the EU, and her opinions about aforementioned Bilderbergers/NWO/international Jewish conspiracy causing the former because of her statements regarding the latter, to form a story where she was hospitalized because she had uncovered the NWO "behind it all". Granted, the bureaucrats and political parties' enthusiasm for the EU has had somewhat of an ugly side, but I still think someone's reaching for a conspiracy theory here.
Most of the links are in Norwegian, but Google Translate does at least passably at translating into English.
Advanced reactors can deal with the waste problem. I like the energy amplifier, which generates the neutrons required for fission externally. That means that there's no chance of a meltdown (just turn off your neutron source) and that the neutrons can also be used to transmute waste into less harmful types. More conventionally, you also have the sodium-cooled fast reactor (basically the IFR, developed further).
From what I gather, using my limited intelligence, you're saying that OLPC-like encrypted, restricted computing platform, could be achieved without the use of DRM?
Yes. Consider again a typical work computer running Windows. On it, the worker usually doesn't have admin rights, and group policy limits what he can do. If this computer had a policy setting saying "only permit the execution of digitally signed files from these companies" (I don't know if Windows actually has that, but bear with me), then the computer will enforce that policy on the user. DRM is not involved because it is possible for the owner of the computer (in this case, the administrator) to override the policy at will; it is not cryptographic magic that keeps non-signed files from being executed, but rather simple explicit privilege limitations.
Maybe this is not really an OLPC-like encrypted restricted computing platform, as the data is not sealed in some vault away from all the users (unlike what is the case for Trusted Computing; I don't know much about the OLPC model itself), but for all intents and purposes, such explicit privilege limitation (policy enforcement) will suffice, except in the case that the owner is the one "you" (whoever you are, then) are trying to limit.
What makes DRM into DRM, as by the Wikipedia definition, is that the "content provider" tries to enforce policy on computers he doesn't own. Remote attestation is just a clever way of getting around this: the content provider only provides the content to computers that lets him enforce policy upon the owner. Now, if we imagined an authoritarian society where all computers are owned by the state and the copyright providers are also part of the state, then that policy enforcement could be done without DRM - simply have the state Admin account enforce the policy upon the "mere users" that lend the computers - but it wouldn't be a very good society.
There is a caveat to this, but one that I think works in the favor of sharing. For most computers, if you have physical access, then you can break the Admin level. But this makes sense, as it's usually the owners who have physical access to the computer, and so it provides a failsafe; if non-owners try to wrest rights from the owner, then the owner can always circumvent them by going up one level. If the OS tries to put the user in a matrix-like hypervisor environment, the user can circumvent the boot process to boot outside of the Matrix, and it can then be cracked at one's leisure.
The flipside of that is that if policies are limited by software, very ingenious children could defeat an OLPC-like lock. That is a social issue, and should be treated as such. Besides, I think that if you're clever enough to be able to do that, you've more or less earned the right anyway. Let the hackers hack; if they do something bad, then tell them. If you really don't want curious hackers, tell them that doing so will mean they'll lose their computers. You can do so because they're not the proper owners. The same goes for escalating privilege to admin on a work machine.
Thus, again to sum up: if you're trying to protect your own computer against others, limited privilege accounts will do nicely. If you're trying to "protect" your own program from being modified on another person's computer, then you can't, and shouldn't, do so.
But I think that this law takes away more freedom than it gives. It does nothing to increase fair use, since you could always your fair use of a public domain work on some other platform. But now, manipulating the bits of a public domain work in a certain way is a crime.
I see it more like the GPL. The public domain is the commons; if you want to publish data provided for common use, then you shouldn't lock it up. The public domain isn't actually GPLed, but the law in question doesn't require you to share your modifications, either; instead, it ensures the rights of distribution inherent in the public domain nature of what is being publis
Can you think of a scenario where blocking DRM legislatively might not be desirable?
Not really. If you want only a select other party to access your data, then you can use ordinary crypto. You're not trying to hide the key from Bob while letting Bob decrypt, you're trying to hide the key from Eve while letting Bob decrypt, something which is much easier to do. (Trusted computing basically boils down to making the computer Bob and the user Eve - the user no longer has root on his own computer)
But let's consider public domain in particular. A work that is in the public domain can be shared freely - or in other words, part of the commons of ideas. There are no laws that act as perimeter fences in idea space around this work (unlike, say, copyrighted works), and so the sharing of that data can only enrich others. "He who receives an idea from me, receives instruction himself without lessening mine; as he who lights his taper at mine, receives light without darkening me", and all that.
If you're dealing with an OLPC situation, then there are two solutions, in my opinion. Either the computer is fully owned by the child in question, in which case the only thing you can do is tell him not to run odd binaries as root (and have a fine-grained security mechanism that limits privilege escalation); or the computer is not, and then the one who really owns it can set a policy that limits the computer to running digitally signed software or what have you. The latter is no more DRM than your average internet kiosk is.
Thus, to sum up, the public domain is already public; artificially limiting works in the public domain is a bad thing as it only constrains sharing without providing any benefits. Restriction policies, on the other hand, can be enforced by the owner but should not be enforced against him.
Slashdot Mirror
They could try replacing all the keys for all devices as a stopgap, but that's pretty problematic and could well just lead to the same leak happening again.
It wouldn't help much in any case, since HDCP is fundamentally flawed. Extracting 40 particular public-private keypairs (perhaps by electron microscopy or what have you) is sufficient to make a synthetic master key.
We can ignore them altogether. Pirate, build heavily encrypted peer-to-peer networks, migrate from the "official" internet to the "dark" internet, until all that's left on the official internet is corporate advertising. Short of a police state, they can't force us to use the official internet, and the more they tighten their grip, the more people find it worthwhile to slip through the fingers into the unofficial internet.
But that's also rather strange. For quantum crypto to make sense, there must be an adversary who can crack Diffie-Hellman (or the key exchange of your choice), but who isn't able to just physically get on the line and insert a man-in-the-middle device. Even if you have perfect quantum crypto, unless you and your intended other party shares a secret, it's impossible to determine if the key negotiation is between you and your intended other party or with Mallory masquerading as that other party, impersonating you to the other party.
That threat model appears to be irrational. Ordinary inside jobs, industrial espionage, whatnot, can't splice in a man-in-the-middle but they can't break DH either - and if quantum computers become prevalent and DH can be easily broken, one can just shift to some post-quantum PK method like McEliece or NTRU.
On the other hand, if the messages are being read by the NSA (who might be able to break DH), quantum crypto won't help, since the NSA could just dig up the cable somewhere and insert a man-in-the-middle device.
That leaves the option where you do have a shared secret. In that case, simply make the shared secret the key for a symmetric key cryptosystem, or if you have the bandwidth, use a one-time pad. QC's only advantage in that case would seem to be that the adversary has no way of stealing the "key" or "pad" without tampering with the cable or compromising one of the endpoints.
(I assume that the communication is being done on a dedicated line, since one can't do quantum crypto over the internet. I also assume that "key exchange" is of the kind where a passive listener has to solve a hard problem to get the key - DH is such a method, but there are others.)
No thinking outside of the box -- no radix sort? Seriously!
In the 2015 version, the bartender shoots first.
We know a man in remote Massachusetts. Every July, peas grow there.
You can't say what would happen in a plurality election based on votes cast in an IRV election. In a plurality with 3 or more candidates, people vote tactically. You can't assume that their single vote would be the same as their first choice in an IRV election.
Alright, but that has no bearing as to the success of AV. If we disregard Plurality, the fact remains that Montroll is both the one-on-one winner and the one with the most first + second place ballots combined. My point was to show that AV is pretty unique in electing Kiss here. Of course you could say that that's not an argument, either, but in the worst case, we can disregard Plurality as it doesn't factor into the IRV/AV vs other methods comparison.
And what does "the worst candidate" mean anyway?
He loses to the other two. If it had been a top two runoff, he would lose no matter who the other candidate had been.
Again, that's crooked thinking. Imagine A is a left wing candidate, but the right has a split vote with candidates B and C both vying for the right. Say that the district is a right wing one. IT may well be that candidate A can individually beat both B and C. But the ideal vote is clearly not to return the left wing candidate for the right wing district. So again you can't just take the IRV votes and munge them into theoretical 2 horse races. It's garbage in, garbage out.
For your first point, consider a left-right spectrum with voters placed along the line and candidates at points close to some of the voters. Then start by "canceling out" the rightmost and the leftmost voter. Do so until there's only one voter left, and the candidate closest to him wins - in other words, the candidate closest to the median voter wins*. If a method satisfies the Condorcet criterion (i.e. picks the candidate that wins all pairups when there is one), then if voters vote honestly (vote candidates closer to them above candidates further out), the candidate closest to the median voter wins. That is called the Black single-peakedness theorem, and implies that a right-wing candidate would win in your example.
If you think about it, it makes sense: if the single left-wing candidate A is to beat both B and C, that means either that a majority puts A ahead of both B and C, in which case it's not a right-wing state, or that some voters rank "B above A above C" or "C above A above B", in which case right versus left wing isn't all that matters.
In a more general multidimensional case, the independence of clones criterion is a good yardstick for whether the method is vulnerable to vote-splitting or not. AV passes it, but so do the advanced Condorcet methods like Schulze. Yee diagrams also show that Condorcet methods find the median voter candidate in 2D instances if the voters are distributed in a Gaussian manner around some central point and the distance metric is Euclidean, although that's a more specific case.
For your second point, note that the pairwise margins are significant. If the voters were aiming to bias AV maximally, the margins would be tighter since the final comparison, when there are only two candidates left, is a pairwise one. Moreover, the fact that the AV opposition was able to unify when AV uniquely elected Kiss, but not in the prior election when it picked the same result as the other methods, suggest that AV did pick the wrong winner - that if there was strategy, it backfired. That is no proof, of course, but it does support the margins conclusion. Finally, we may add the first+second vote metric to the heap; if there was strategy, it would be unlikely that the strategy would show a clear winner by both measures, because if each voter crafts his vote to maximally affect the AV outcome, there would be few degrees of freedom left to make someone else the Condorcet winner and Bucklin (first plus second) winner as well.
* This is a reasonable g
But you've come up with a non-realistic, very carefully worked out example that appears to do the wrong thing, ever so slightly. But every electoral system is capable of being criticised from some such carefully worked out example. With FPTP you don't even need to carefully work out an example. Remember that FPTP regularly elects people with only 35%ish of the vote.
Unless you live in a place where ballot stuffing is a problem*, AV is indeed better than FPTP. It is wrong, however, to say that AV's failures only occur in contrived examples. AV works as long as the minor parties are so small that they get eliminated before any major parties, but when that is not the case, the parties may be eliminated in the wrong order. Here's an example of that, from the 2009 Burlington, Vermont election for Mayor:
1332 voted Montroll > Kiss > Wright
767 voted Montroll > Wright > Kiss
455 voted Montroll
2043 voted Kiss > Montroll > Wright
371 voted Kiss > Wright > Montroll
568 voted Kiss
1513 voted Wright > Montroll > Kiss
495 voted Wright > Kiss > Montroll
1289 voted Wright
Plurality elects Wright, the worst candidate of the three. IRV elects Kiss even though in a real runoff, Montroll could beat any of the others one-on-one and had more first plus second place preferences than either of the others. The reason why that happened was because other first preference votes obscured Montroll's support and so led to him being eliminated. Because IRV (AV) misjudged Kiss's support and elected him winner, the voters that thought that was a bad idea were able to unite and later replace IRV with ordinary top-two runoff.
* The reason for this is that FPTP can be summed locally: just count first preferences and send the array to the central location. AV has to be counted centrally so the actual ballots have to move further, unless you have the required infrastructure in place to send messages back and forth about whom to eliminate at each stage.
So party A has the most first choice votes, and party A has the most second choice votes. But party B gets in. Instead of making 49% of people completely happy and 48% slightly happy, you're making 48% completely happy and 33% slightly happy. Why are you giving the final say to the second choice of those who have voted for the least popular candidates?
That's because AV is not a very good single-winner method. What you want is something more like what Wikimedia uses - a Condorcet method, where each candidate is counted as beating the candidates ranked below it, and the candidate that beats every other one-on-one (like in sports) wins. Unfortunately, it's too radical (with a very few exceptions, no such method has been used for governmental elections) and so it has absolutely no chance even in situations where using a single-winner method would make sense (like electing a president or a party leader).
For your example, a simple count-the-winning-side Condorcet method would give:
A preferred to B by 49, B preferred to A by 51, B wins and gets 51 points
A preferred to C by 97, C preferred to A by 2, A wins and gets 97 points
A preferred to D by 97, D preferred to A by 1, A wins and gets 97 points
B preferred to C by 69, C preferred to B by 22, B wins and gets 69 points
B preferred to D by 70, D preferred to B by 10, B wins and gets 70 points
C preferred to D by 22, D preferred to C by 10, C wins and gets 22 points
and the outcome is: A: 194 pts, B: 190 pts, C: 22 pts, D: nil.
There are better systems (Wikimedia uses the Schulze method), but they are also more complex.
That breaks the fundamental principal of a representative being the voice of his district. If the people living in Town A vote for a Liberal, and he gets replaced by this new AV+ system by a smaller psrty (say libertarian) then he's not really representing that town's views.
True. I would prefer STV with 3-5 member districts, myself, but that wasn't even on the agenda.
Now, you could say that multimember districts like those STV require would weaken the principle of representation because each district would be larger, but that's part of a necessary tradeoff. In order to give minorities more than a zero percent say, one has to weaken the majorities; but at least with STV, the later preferences of those whose candidates didn't win can make a difference. The 50%+/-1 example shows why: in order to give the 50%-1 minority some say, some candidates that would otherwise be allocated to the 50%+1 majority has to go to the 50%-1 minority instead, and there's no way of getting around that. MMP does the tradeoff very abruptly: the constituency part of the parliament is completely majority-based and then the list part patches it up - no majority loss in the constituency part, great loss in the list part. STV does the tradeoff more subtly: a little bit in each district in order to permit PR at all.
If the promise to have a referendum on Alternative Voting is delivered upon, and the electorate are intelligent enough to vote it in, then it will solve this predicament. It will make it always advantageous to vote for the party(s) you prefer, rather than voting tactically for a different party in the hope of keeping the villain of choice out.
AV provides slightly more fair rules, but not enough. To show this most clearly, imagine there are only two parties, and one of the parties gets 50% + 1 of every single constituency. Then half the voters' votes are wasted. A much more fair outcome would give half the seats to the second party, and for that you'll need STV or party list. The Liberal Democrats wanted AV+, which is a combination of AV and party list wherein a party that gets too few constituency seats is awarded top-up seats to compensate. However, AV+ lost in the compromise because the Conservatives don't want proportional representation, and thus they arrived at plain old AV. As Australia shows, it's not enough: Australia uses AV and has a two plus a half party system (Labor on the one hand and National plus Liberal on the other), even with the Senate, which uses proper PR, to counterbalance it.
But if the AV referendum passes, perhaps it will lead to another about true PR. On the other hand, it could also become a reform without reform, discouraging voters and parties from considering better systems because "we already tried that and it didn't help".
The difference is that on the internet, you can use pseudonymity. You can call yourself JSBiff and I can call myself kvezach, and neither of us will know the other person's real name. My different pseudonyms can be limited from each other, as well: if I register as foo on a chess site and bar on a Go site and say the other game sucks on both, nobody will call me on it (if I don't do anything stupid).
In contrast, DRM fails because once you can see it, you can copy it, no matter what pseudonym you (or the creator) may be hiding under. The analogous situation would be that stream-games-from-server thing that has been on Slashdot before: it only shows you what it decides you need to see, so you can't copy the game itself. Well, to stretch the analogy, it would be like a peer-to-peer network version of that, where you don't even know where the game is ultimately stored (i.e. pseudonymity).
The net effect of which would be ... letting others decide for you. The same as voting over the line, just in a slightly less predictable fashion. I don't get why it should make such an important difference?
It makes a difference because you can stop at exactly the point you want. To use an American example: optional preference voting would let you rank R, D, and Green candidates without having to care about where you're going to put the Constitution Party. Or a Norwegian one: optional preference voting would let you rank AP relative to FrP without having to care about KSP or NKP. The way Australia's "preference" voting currently works, you would have to rank even the obscure candidates. That's a lot of work, so most people just vote above the line, and so the fine-grained multiwinner system (that has the potential to be proportional not just by party) is reduced to, in essence, a very opaque party list method.
Unless you use old ActiveX programs that don't support newer versions of IE, that is.
The seventh one is fixed, so we have 38 symbols for 6: 38*37*36*35*34*33 ~= 2*10^9, ends up being 3779 years. That's still too much, but...
There are these pretty crystals inside of the slanted table with all the buttons on it though.
So you have a DHD! If you remember, putting in the wrong address on a DHD just blinks the buttons and nothing happens. Therefore, the activation sequence takes much shorter time than a minute. Let's say that you could get it down to a second with the proper robotics. Then you could exhaust the space in 63 years. Also, the DHD only has 38 symbols, and you can extract about 50 addresses from each if you know how. If you could figure out how, you could make a robot that steals the addresses, tries them all, and repeats for those that actually open.
You have to get to 70% of c just to make a 500-ly journey feel like 500 years, and even closer if you want it to feel quicker. Good luck getting that close to c without antimatter. Good luck getting that close to c even with antimatter.
I can turn it on, but I haven't found a combination of these freakin' buttons to do anything useful.
JW, this is David Lightman. David, this is JW. Well, David, why don' you tell JW about this amazing trick you've discovered on your computer? I think you're calling it "wardialing" or something similar...
I remember a story a few years ago saying congressional approval levels are always dismally low and yet historically about 90% of congressmen are re-elected. Why? Because the majority of people absolutely hate Congress but think their personal representatives are doing a pretty good job.
And perhaps also gerrymandering. When the representatives can pick their voters before the voters vote, then it's no wonder the incumbency rate is so high. Contorted districts like this one aren't drawn just for the amusement of the mapmakers.
Great idea! Why not spend money on a warp drive while we're at it?
It does seem that they were very focused on being able to extract the oil rather than just stopping the leak. Now, I'm not an engineer, but could their desire for continued extraction of oil have delayed their plans, made the stack more complex?
In any case, we'll see whether it works. Hopefully it'll at least buy them enough time to drill relief wells.
really, there's no way around this that can't also be worked around by the spammers. Every single step is met by counter action and evasion. The only thing that works is jail time.
How about the idea of proof-of-work with price discrimination? Unknown domains start at a fairly high level, so it takes a long time to send mail. If the user or domain has sent a few good mails, it's "trusted" (but using a robust trust metric so that spammers can't just trust each other), and then sending mail is fast. If the user gets compromised, he drops on the trust network and suddenly it's expensive to send mail again.
It's kinda a hack because you're explicitly wasting processing (or memory access) power to limit something that's otherwise unlimited, but it could work. That is, it could work once someone waves his wand and establishes that distributed trust network.
Botnets would initially circumvent this, but the zombies would get their trust degraded and so would send spam very slowly, making them unsuitable. Thus the spammers would have to ever chase new hosts. The delay function can be structured so that "stop and go" (send some spam until at worst trust, then let the user get back to good trust, then send more spam) doesn't work.
The closest thing I could find that would fit his description is this, but you have to cut through a lot of noise about how the Bilderbergers/NWO/international Jewish conspiracy is controlling the world to get at what it's actually saying. Basically, it seems to have mixed up a person's forcible hospitalization, her statements against the incorporation of Norway into the EU, and her opinions about aforementioned Bilderbergers/NWO/international Jewish conspiracy causing the former because of her statements regarding the latter, to form a story where she was hospitalized because she had uncovered the NWO "behind it all". Granted, the bureaucrats and political parties' enthusiasm for the EU has had somewhat of an ugly side, but I still think someone's reaching for a conspiracy theory here.
Most of the links are in Norwegian, but Google Translate does at least passably at translating into English.
Advanced reactors can deal with the waste problem. I like the energy amplifier, which generates the neutrons required for fission externally. That means that there's no chance of a meltdown (just turn off your neutron source) and that the neutrons can also be used to transmute waste into less harmful types. More conventionally, you also have the sodium-cooled fast reactor (basically the IFR, developed further).
From what I gather, using my limited intelligence, you're saying that OLPC-like encrypted, restricted computing platform, could be achieved without the use of DRM?
Yes. Consider again a typical work computer running Windows. On it, the worker usually doesn't have admin rights, and group policy limits what he can do. If this computer had a policy setting saying "only permit the execution of digitally signed files from these companies" (I don't know if Windows actually has that, but bear with me), then the computer will enforce that policy on the user. DRM is not involved because it is possible for the owner of the computer (in this case, the administrator) to override the policy at will; it is not cryptographic magic that keeps non-signed files from being executed, but rather simple explicit privilege limitations.
Maybe this is not really an OLPC-like encrypted restricted computing platform, as the data is not sealed in some vault away from all the users (unlike what is the case for Trusted Computing; I don't know much about the OLPC model itself), but for all intents and purposes, such explicit privilege limitation (policy enforcement) will suffice, except in the case that the owner is the one "you" (whoever you are, then) are trying to limit.
What makes DRM into DRM, as by the Wikipedia definition, is that the "content provider" tries to enforce policy on computers he doesn't own. Remote attestation is just a clever way of getting around this: the content provider only provides the content to computers that lets him enforce policy upon the owner. Now, if we imagined an authoritarian society where all computers are owned by the state and the copyright providers are also part of the state, then that policy enforcement could be done without DRM - simply have the state Admin account enforce the policy upon the "mere users" that lend the computers - but it wouldn't be a very good society.
There is a caveat to this, but one that I think works in the favor of sharing. For most computers, if you have physical access, then you can break the Admin level. But this makes sense, as it's usually the owners who have physical access to the computer, and so it provides a failsafe; if non-owners try to wrest rights from the owner, then the owner can always circumvent them by going up one level. If the OS tries to put the user in a matrix-like hypervisor environment, the user can circumvent the boot process to boot outside of the Matrix, and it can then be cracked at one's leisure.
The flipside of that is that if policies are limited by software, very ingenious children could defeat an OLPC-like lock. That is a social issue, and should be treated as such. Besides, I think that if you're clever enough to be able to do that, you've more or less earned the right anyway. Let the hackers hack; if they do something bad, then tell them. If you really don't want curious hackers, tell them that doing so will mean they'll lose their computers. You can do so because they're not the proper owners. The same goes for escalating privilege to admin on a work machine.
Thus, again to sum up: if you're trying to protect your own computer against others, limited privilege accounts will do nicely. If you're trying to "protect" your own program from being modified on another person's computer, then you can't, and shouldn't, do so.
But I think that this law takes away more freedom than it gives. It does nothing to increase fair use, since you could always your fair use of a public domain work on some other platform. But now, manipulating the bits of a public domain work in a certain way is a crime.
I see it more like the GPL. The public domain is the commons; if you want to publish data provided for common use, then you shouldn't lock it up. The public domain isn't actually GPLed, but the law in question doesn't require you to share your modifications, either; instead, it ensures the rights of distribution inherent in the public domain nature of what is being publis
Can you think of a scenario where blocking DRM legislatively might not be desirable?
Not really. If you want only a select other party to access your data, then you can use ordinary crypto. You're not trying to hide the key from Bob while letting Bob decrypt, you're trying to hide the key from Eve while letting Bob decrypt, something which is much easier to do. (Trusted computing basically boils down to making the computer Bob and the user Eve - the user no longer has root on his own computer)
But let's consider public domain in particular. A work that is in the public domain can be shared freely - or in other words, part of the commons of ideas. There are no laws that act as perimeter fences in idea space around this work (unlike, say, copyrighted works), and so the sharing of that data can only enrich others. "He who receives an idea from me, receives instruction himself without lessening mine; as he who lights his taper at mine, receives light without darkening me", and all that.
If you're dealing with an OLPC situation, then there are two solutions, in my opinion. Either the computer is fully owned by the child in question, in which case the only thing you can do is tell him not to run odd binaries as root (and have a fine-grained security mechanism that limits privilege escalation); or the computer is not, and then the one who really owns it can set a policy that limits the computer to running digitally signed software or what have you. The latter is no more DRM than your average internet kiosk is.
Thus, to sum up, the public domain is already public; artificially limiting works in the public domain is a bad thing as it only constrains sharing without providing any benefits. Restriction policies, on the other hand, can be enforced by the owner but should not be enforced against him.