100% have the physical ability to consume their fat. Calories out minus calories in - the balance will come from somewhere. Claiming anything else is to claim that the human body is a perpetual motion machine. What they can't is seem to accomplish is to develop an objective sense of what they need; I know one fat person who tells me he needs some ridiculous amount of protein every day or he's going to wither away. He thinks he needs 3000 kcal per day or he's going to starve - then tells me he's fat because of his genes. Last I checked there was no 'clue' gene.
It's definitely a race to the bottom, because most parents have no clue what they buy their kids. The #1 goal of buying something is to shut the kid up. The #2 goal is to surprise the kid with a gift. And a $10 CrapKit will do either just as well as a $50 quality one. Toys are considered disposable. And the kid has no clue about concepts like quality and functionality - as long as it looks the part. (Brand recognition is a factor.) The kid will play with the CrapKit, find it difficult to proceed beyond the basics, and will likely grow tired of it because of its limitations. The parent will observe that the hotly desired toy stops being played with after a few days or weeks and pats themselves on the back for being cheap and wonders why anyone would buy the expensive version. They leave this to the people with more money than sense. It's a self-reinforcing spiral, simply because the average person is average intelligence, which means if you even bring up the subject of developing intelligence they'll look at you like you're from a different planet. It's just not something that they ever spend a single brain cycle on. Hey, they came out alright... right?
I always played with Legos when I was a kid. Well, to my parents it was playing, to me it was construction projects. As I got older they became ever better planned and thought out, and I'd carefully plan around the parts available. My parents never saw that part. They viewed Lego, I'm sure, as the equivalent of a crayon and a sheet of paper.
By the time I was 10 or so I built things like flexible suspension bridges (suspended with string) that could carry my HO size train set across 3 feet or more, to replicas of buildings I read about. Lego is a fantastic tool for early development of an innate sense of force distribution; in particular how to design for forces to distribute into compression with little pulling (depending on axis) and close to zero twisting. It encourages focusing on difficult problems somewhat beyond the current skill, then learning through failure and developing an innate sense for how to further improve something that a bystander might already be impressed by or think is beyond good enough.
My take on it is that every parent should buy their kid real, quality Lego. Mostly generic blocks. At least give it a try. Because if the kid takes to it - boy are you getting something of real value for dirt cheap!
IPsec is unusable for this since it also hides the headers. Firewalls need to see headers to enforce policy, and if they can't they'll rightfully reject traffic. (I.e., reject what can't be positively confirmed as acceptable.) With an open key distribution scheme as proposed in obstcp (via DNS CNAME) a firewall that lets IPsec through might as well not exist. This is one of the big reasons IPsec today is used for VPN tunnels between firewalls/routers and just about nothing else. An ISP that sees IPsec can safely assume it's a VPN tunnel and you're telecommuting.
Under the proposed key distribution scheme, a man-in-the-middle relay attack requires DNS hijacking or spoofing as well. Sure, it can be done. But this raises the bar for an ISP from "sure, just listen in!" to the legally risky. Do you think Comcast would spoof DNS for, say, a speakeasy customer? Or obs.example.com? I doubt they'd dare touch it, and if they did they'd soon find themselves in court.
The problem with the RS/Heath etc kits is that they still didn't explain how anything worked. It was still just a connect-the-dots exercise, except with 'real' components and a hot soldering iron.
And herein lies the beef: the skill of being able to populate a printed board is orthogonal to understanding how any of it works.
I'd recommend a simple breadboard kit, some tools (wire stripper, cutter), components, simple discrete logic (if you're interested in digital), 5V and 12V DC supplies, a DMM, maybe a cheap scope, etc. There are many places to get it; jameco.com for instance. Mouser is good too. Jameco also sells educational kits, but I'd still recommend a breadboard setup and a box of discrete parts for experimentation; educational kits can be useful teaching aids, but without the teaching parts they're not of that much use IMO.
Of course, a good text book intro to electronics is required as well. Preferably one with exercises you can recreate on the breadboard.
I created a 4 hour video on building boats. I sell probably 30 a year. An online video rental site bought a legal copy and now rents my video out at $15 per week. So why aren't you renting out your video yourself for $15/week? Then if people want a hardcopy offer to sell it for $25 extra or $40 without a prior rental. It won't eliminate the possibility of a middle man renter, but it makes you a more attractive source. It's probably a better business model, too.
Or, you can go threaten the middle man renter with switching to this business model, but offer them to carry out the sales and send you a check. (Minus their commission.) They'd likely be happy to get on your good side and expand their business further.
And for heaven's sake, make it a DVD. Selling tapes is probably your #1 problem - so many people just don't have tape players anymore and they go rent a download just so they can play it at all.
Anything within the event horizon of the hole, by definition, cannot escape to the outside universe again.
But it can still be observed in some ways, no? Matter inside the horizon exerts gravitational pull on matter outside - so matter inside the horizon still in the process of accreting should produce gravitational effects that can be observed in the matter still outside. We should be able to observe what goes on inside, in some ways. Similarly, a star cluster getting sucked in should be able to cause measurable wobble in the location, and some sort of tidal effects inside?
My rather layperson understanding is that the findings prove there are N-order effects in evolution. Given what we know about the complex interaction of genes and how they switch one and other off in complex networks, there are many layers of order where changes can occur, and conversely any one change could impart both, say, a bigger eye as well as a tendency to evolve say the skin in some direction. So you can have one immediately beneficial change, like a slightly tweaked eye, that takes hold quickly act to set up the species for other future directional changes -- or even 2-, 3-, or N-th order changes (like changes to the switching graph itself).
Evolution also isn't centered around individual procreation. Clearly few ants or bees procreate, yet they are extremely successful as species. Evolution is about the success of the species, and can't be reduced to a 'fittest of the lineages' even in species where all or most individuals are genetically enabled to do so.
Anyway, I'm sure a real geneticist or other professional in the field could really clue us in much better.
With less stuff for our immune systems to fight, autoimmune diseases are more common. Here's a link to check out, just for one reference.
The link refers to parasites deactivating or suppressing the immune system, and that this reduces auto-immune response. Just like drugs against the same illnesses, which also suppress the auto response.
This is very different from stating that auto-immune illnesses are caused by insufficient exposure in general, or that someone without such an illness would benefit from suppression therapy. (Or parasites, for that matter.)
You could just as well argue that because of low infant mortality (due to excess hygiene and antibiotic therapies?) people with defective immune systems are more likely to survive long enough to talk about it and procreate. There's no proof for this either.
just go sit down in a coffee shop and relax, read the paper, without ordering anything?
Yes dammit, this should be a felony too! More government! More police! More laws! God damn criminals are ruining our moral standards! Make everything a felony! (Except vehicular laws, I want to be able to drive any way I please.)
Ten miles isn't really considered a commute in the SF Bay Area. 50-100 mi is more normal. I did it for 14 years up and down the peninsula (SF to SV). Now my commute is only 6mi since I work in the city, and yes, I ride a bicycle. I don't bother riding on the occasional rainy day though, I'll just hop on Muni since I live by a trolley stop and my job is at the other end.
It's not categorized as wiretapping in Sweden. There, anyone can receive any radio frequency, you only need a license to transmit. The ether is considered a public space. This applies both to individuals and the government, all that's needed is a suitable receiver to listen in.
Another affirm, I've seen this behavior for YEARS. Spammers will try ALL the MX's for a domain and not just because the primary "failed".
Yeah. My first thought when seeing this was wondering why he'd make the non-responder the primary, I'd make it the second out of three!
This plays better with sendmail too (which is what I use), which configures itself automatically as primary/secondary/neither by looking at the MX records. If I list a TCP SYN hole as my primary my real primary will think it's a secondary and simply hold mail indefinitely. But making the first secondary a non-responder might be useful. Simply have it be an unused address, that way the spammer will wait for timeout. (No doubt shortened so they can move on, but still a waste of their attack window.)
This makes me wonder if there's not an easier way to bug spammers... Add a TCP sockopt to delay the SYN ACK for a specific listen socket by some time period, like 15 seconds. For a legitimate SMTP connection this is a non-issue, for a spammer it's a tremendous delay. Might add a bit of kernel state for a very busy mail server though. The first time through, the tuple is greylisted to bypass the delay next time. If, however, the other party times out and responds to the SYN ACK with a RST, or sends it unsolicited, it's on an impractically short timeout and can probably be blacklisted as a spammer.
ISPs not only should be rate-limited, but if people try to exceed the limit, they should check if it was spamming and block whatever customer is at that IP.
Rate limiting is very useful, but some customers actually do need to send bulk email. They run mailing lists, have regular newsletters, etc. However, there's a strong correlation between customers who have no intention of ever sending bulk email and customers who get zombied. So the automatic block is a good idea, followed by instructions sent to the customer on how to de-zombify their PC. No instructions on how to bypass the rate limiter. A customer who actually needs to send bulk mail will contact customer support, in person, to get themselves whitelisted. That way the ISP will 1) detect zombies, 2) have a handle on which of their customers are legitimate bulk mailers. To defeat rate limiting spammers will have to leave a lot more tracks, making it less attractive.
He's certainly signed a contract, or he wouldn't be on the payroll! Not only that, but law enforcement officers are a huge legal special case with many of their rights curtailed. They are, for instance, LEOs 24hrs a day whether in uniform or not. It's not a 9-to-5 job with well defined boundaries. In addition, many states have statutory laws making intellectual property developed on the public dime public domain. Which this should be. All he can reasonably demand is overtime pay for the off-duty hours. But an employer can't be expected to lose the output of an employee just because the latter for whatever reason didn't file overtime.
CW *is* useful though, and I've come to embrace it for
the VHF/UHF weak signal stuff I've been doing, where at
time the luxury of a voice just isn't there; things are
too weak.
If you're going to send messages, which is probably what you'll want at low bandwidths, there's got to be better and more efficient encodings and transmission protocols than CW. Off-hand, how about not sending the message in order so transmission errors don't result in consecutive symbols lost, and with CRC/ECC techniques and encapsulation to boost the chances of recovery (and reduce sensitivity to noise/loss). Isn't this the kind of thing that makes experimenting with 'moon bounce' and such fun in the first place? Experimenting with encodings, compressions, recovery methods, heuristics, homing algorithms, etc?
If I were part of the ham culture I'd be concerned that CW, by providing a predesigned but rather poor encoding and protocol standard, discourages innovation and entrenches mediocrity.
Solar cells cost a lot of energy to make, so what's the life span on these things? What's left if you subtract the manufactoring costs from the life-time energy generation of these things?
The info might not be the latest, but scroll down to the section on Energy Payback Time (EPBT): http://www1.eere.energy.gov/solar/pv_basics.html. Looks like a couple of years, give or take. If the technology with the 40% efficiency mentioned is equally cost-effective, then we're probably down under a year.
The tally should be done at the optical scanner that scans in the printed ballots. The optical scanner/tabulator software of course should be bulletproof and not easily modifiable.
This just moves the burden of tamper-proofing from the machines in the booths to the one machine that scans the slips. If you can't tamper-proof the former, then you can't tamper-proof the latter either. Also, "not easily modifiable" isn't good enough, and is actually orthogonal to the issue; the question is not how hard it is to tamper with, but how hard it is to detect that it's been compromised. The equipment needs to be designed so that any compromise is obvious. Checkboxes on a printed card works quite nicely for this, it's perfectly clear whether a candidate has been removed from some ballots for instance. Any electronic machine needs to be equally obvious to the voter when it has been tampered with. I personally doubt this is even possible, and suspect it's the wrong approach.
Since Diebold has a crappy track record with electronic voting, why should we as consumers have any confidence in their ATMs?
We trust ATMs because they are part of the banking system, and banks are trust corporations. Trust is their #1 product. If we didn't trust them we'd stash our money in our mattresses. People who don't trust banks do.
This is very different from a political system, where nobody trusts (and shouldn't trust) anybody else. The voter shouldn't trust the poll staff or equipment vendor, the poll staff shouldn't trust the voter or equipment vendor, etc. Basically, it's a system of mutual distrust. Such a system can only function on the principle of transparency and mutual verification. (Like Reagan said about the USSR, "trust but verify.") And this is where the electronic voting machine is incompatible: it's too complex and too concealed. There is no way for a voter to look and say yeah, the firmware looks trustworthy on this one. As opposed to punching a hole in a ballot, where it's pretty obvious whether the right hole got punched; the punch lever isn't going to jump and punch some other predetermined hole based on a statistical distribution, it's not complex enough to permit that degree of manipulation. In fact, the very act of bringing an unverifiable, complex machine into the center of the election process may be rooted in ignorance, but needs to be treated as a trojan horse. Unless it can be verified otherwise, an assumption needs to be made that it's compromised.
Elections is one place where 'simple and mechanical' is inherently better!
I bet that if a senior engineering class at Purdue (not even MIT) put their minds to the task they could build a mechanical voting machine that would not create the same level of controversy as Diebold's machines.
The problem isn't one of engineering, but of trust. Or, rather, mutual distrust. A system like that (e.g. election system) can't function in the long term without transparency and verification by all parties. This means the system needs to be a simple as possible, preferably with mechanical parts that can be eyeballed by anyone familiar with the system. After verification, the party (political parties, election officials, equipment maker) needs to be able to apply their seal to it. A firmware or electronic system in general is inherently unsuitable because of the difficulty of verification. It's also very difficult to adequately tamper proof, while DRM can be used it's not easy to check whether some other party has replaced the DRM system itself with a lookalike.
Another system built around mutual distrust was the cold war relationship between the USSR and the USA. Both parties accepted a high degree of spying and snooping, because it created some level of transparency and permitted mutual verification. Without rampant spying chances are good the paranoia had grown on both sides until it reached a point of war. 'Trust but verify'.
A lot of the cable/satellite HD isn't really HD (1080i) but lower resolution at as low as 4-6Mbps. Same bitrate as DVD, so don't be too surprised if it doesn't look much better. By HD they usually mean "a little better than DVD", but it's not a huge difference. Often it's worse that plain old DVD. It also varies from show to show, I've found the local cable HD sports broadcasts to be pretty good, but still not HD. I don't have a set yet, because I can't see the benefit either. I just looked around to see what's out there and wasn't impressed. The only way most of us are going to get HD content is through a player, so I'm simply sitting it out until there is one established format -- at which point 1080i sets should be more common, far less expensive, and demonstrably usable to their fullest with the players and STBs.
If the DRM software is all open source (ie it doesnt need to be linked against closed source libraries to compile which actually contains the DRM parts), how can it prevent me from disabling the license checks for burning the media to disk,
The software never sees content in plaintext, that would be utterly pointless.
A conditional access system needs to do all the backend work: interact with the subscriber and device databases, handle purchase requests, produce billing and audit data, look for fraud, revoke keys, assist with customer service, produce a gazillion different business reports, and of course issue keys. It needs to be tamper proof and self contained with a minimum of external interfaces. This is the space Sun has made an announcement in. They already had a framework to deliver the content plus the necessary client side support. On the client device that is requesting keys, managing keys, picking the right key out of a key block, and stuffing keys into or providing them to the secure decoder as needed. The delivery part is download or streaming.
Slashdot Mirror
100% have the physical ability to consume their fat. Calories out minus calories in - the balance will come from somewhere. Claiming anything else is to claim that the human body is a perpetual motion machine. What they can't is seem to accomplish is to develop an objective sense of what they need; I know one fat person who tells me he needs some ridiculous amount of protein every day or he's going to wither away. He thinks he needs 3000 kcal per day or he's going to starve - then tells me he's fat because of his genes. Last I checked there was no 'clue' gene.
It's definitely a race to the bottom, because most parents have no clue what they buy their kids. The #1 goal of buying something is to shut the kid up. The #2 goal is to surprise the kid with a gift. And a $10 CrapKit will do either just as well as a $50 quality one. Toys are considered disposable. And the kid has no clue about concepts like quality and functionality - as long as it looks the part. (Brand recognition is a factor.) The kid will play with the CrapKit, find it difficult to proceed beyond the basics, and will likely grow tired of it because of its limitations. The parent will observe that the hotly desired toy stops being played with after a few days or weeks and pats themselves on the back for being cheap and wonders why anyone would buy the expensive version. They leave this to the people with more money than sense. It's a self-reinforcing spiral, simply because the average person is average intelligence, which means if you even bring up the subject of developing intelligence they'll look at you like you're from a different planet. It's just not something that they ever spend a single brain cycle on. Hey, they came out alright... right?
I always played with Legos when I was a kid. Well, to my parents it was playing, to me it was construction projects. As I got older they became ever better planned and thought out, and I'd carefully plan around the parts available. My parents never saw that part. They viewed Lego, I'm sure, as the equivalent of a crayon and a sheet of paper.
By the time I was 10 or so I built things like flexible suspension bridges (suspended with string) that could carry my HO size train set across 3 feet or more, to replicas of buildings I read about. Lego is a fantastic tool for early development of an innate sense of force distribution; in particular how to design for forces to distribute into compression with little pulling (depending on axis) and close to zero twisting. It encourages focusing on difficult problems somewhat beyond the current skill, then learning through failure and developing an innate sense for how to further improve something that a bystander might already be impressed by or think is beyond good enough.
My take on it is that every parent should buy their kid real, quality Lego. Mostly generic blocks. At least give it a try. Because if the kid takes to it - boy are you getting something of real value for dirt cheap!
IPsec is unusable for this since it also hides the headers. Firewalls need to see headers to enforce policy, and if they can't they'll rightfully reject traffic. (I.e., reject what can't be positively confirmed as acceptable.) With an open key distribution scheme as proposed in obstcp (via DNS CNAME) a firewall that lets IPsec through might as well not exist. This is one of the big reasons IPsec today is used for VPN tunnels between firewalls/routers and just about nothing else. An ISP that sees IPsec can safely assume it's a VPN tunnel and you're telecommuting.
Under the proposed key distribution scheme, a man-in-the-middle relay attack requires DNS hijacking or spoofing as well. Sure, it can be done. But this raises the bar for an ISP from "sure, just listen in!" to the legally risky. Do you think Comcast would spoof DNS for, say, a speakeasy customer? Or obs.example.com? I doubt they'd dare touch it, and if they did they'd soon find themselves in court.
You can more simply boycott them if they're required to disclose that their products are DRM infested and come with restrictive install policies.
The problem with the RS/Heath etc kits is that they still didn't explain how anything worked. It was still just a connect-the-dots exercise, except with 'real' components and a hot soldering iron.
And herein lies the beef: the skill of being able to populate a printed board is orthogonal to understanding how any of it works.
I'd recommend a simple breadboard kit, some tools (wire stripper, cutter), components, simple discrete logic (if you're interested in digital), 5V and 12V DC supplies, a DMM, maybe a cheap scope, etc. There are many places to get it; jameco.com for instance. Mouser is good too. Jameco also sells educational kits, but I'd still recommend a breadboard setup and a box of discrete parts for experimentation; educational kits can be useful teaching aids, but without the teaching parts they're not of that much use IMO.
Of course, a good text book intro to electronics is required as well. Preferably one with exercises you can recreate on the breadboard.
Or, you can go threaten the middle man renter with switching to this business model, but offer them to carry out the sales and send you a check. (Minus their commission.) They'd likely be happy to get on your good side and expand their business further.
And for heaven's sake, make it a DVD. Selling tapes is probably your #1 problem - so many people just don't have tape players anymore and they go rent a download just so they can play it at all.
Anything within the event horizon of the hole, by definition, cannot escape to the outside universe again.
But it can still be observed in some ways, no? Matter inside the horizon exerts gravitational pull on matter outside - so matter inside the horizon still in the process of accreting should produce gravitational effects that can be observed in the matter still outside. We should be able to observe what goes on inside, in some ways. Similarly, a star cluster getting sucked in should be able to cause measurable wobble in the location, and some sort of tidal effects inside?
My rather layperson understanding is that the findings prove there are N-order effects in evolution. Given what we know about the complex interaction of genes and how they switch one and other off in complex networks, there are many layers of order where changes can occur, and conversely any one change could impart both, say, a bigger eye as well as a tendency to evolve say the skin in some direction. So you can have one immediately beneficial change, like a slightly tweaked eye, that takes hold quickly act to set up the species for other future directional changes -- or even 2-, 3-, or N-th order changes (like changes to the switching graph itself). Evolution also isn't centered around individual procreation. Clearly few ants or bees procreate, yet they are extremely successful as species. Evolution is about the success of the species, and can't be reduced to a 'fittest of the lineages' even in species where all or most individuals are genetically enabled to do so. Anyway, I'm sure a real geneticist or other professional in the field could really clue us in much better.
With less stuff for our immune systems to fight, autoimmune diseases are more common. Here's a link to check out, just for one reference.
The link refers to parasites deactivating or suppressing the immune system, and that this reduces auto-immune response. Just like drugs against the same illnesses, which also suppress the auto response.
This is very different from stating that auto-immune illnesses are caused by insufficient exposure in general, or that someone without such an illness would benefit from suppression therapy. (Or parasites, for that matter.)
You could just as well argue that because of low infant mortality (due to excess hygiene and antibiotic therapies?) people with defective immune systems are more likely to survive long enough to talk about it and procreate. There's no proof for this either.
just go sit down in a coffee shop and relax, read the paper, without ordering anything?
Yes dammit, this should be a felony too! More government! More police! More laws! God damn criminals are ruining our moral standards! Make everything a felony! (Except vehicular laws, I want to be able to drive any way I please.)
It means 60 out of 100,000 were able to conclusively prove their innocence.
Ten miles isn't really considered a commute in the SF Bay Area. 50-100 mi is more normal. I did it for 14 years up and down the peninsula (SF to SV). Now my commute is only 6mi since I work in the city, and yes, I ride a bicycle. I don't bother riding on the occasional rainy day though, I'll just hop on Muni since I live by a trolley stop and my job is at the other end.
It's not categorized as wiretapping in Sweden. There, anyone can receive any radio frequency, you only need a license to transmit. The ether is considered a public space. This applies both to individuals and the government, all that's needed is a suitable receiver to listen in.
I have a suggestion: contact BT customer support and they'll help you resolve this. It's a well-understood WMP settings issue with a simple fix.
Yeah. My first thought when seeing this was wondering why he'd make the non-responder the primary, I'd make it the second out of three!
This plays better with sendmail too (which is what I use), which configures itself automatically as primary/secondary/neither by looking at the MX records. If I list a TCP SYN hole as my primary my real primary will think it's a secondary and simply hold mail indefinitely. But making the first secondary a non-responder might be useful. Simply have it be an unused address, that way the spammer will wait for timeout. (No doubt shortened so they can move on, but still a waste of their attack window.)
This makes me wonder if there's not an easier way to bug spammers... Add a TCP sockopt to delay the SYN ACK for a specific listen socket by some time period, like 15 seconds. For a legitimate SMTP connection this is a non-issue, for a spammer it's a tremendous delay. Might add a bit of kernel state for a very busy mail server though. The first time through, the tuple is greylisted to bypass the delay next time. If, however, the other party times out and responds to the SYN ACK with a RST, or sends it unsolicited, it's on an impractically short timeout and can probably be blacklisted as a spammer.
Rate limiting is very useful, but some customers actually do need to send bulk email. They run mailing lists, have regular newsletters, etc. However, there's a strong correlation between customers who have no intention of ever sending bulk email and customers who get zombied. So the automatic block is a good idea, followed by instructions sent to the customer on how to de-zombify their PC. No instructions on how to bypass the rate limiter. A customer who actually needs to send bulk mail will contact customer support, in person, to get themselves whitelisted. That way the ISP will 1) detect zombies, 2) have a handle on which of their customers are legitimate bulk mailers. To defeat rate limiting spammers will have to leave a lot more tracks, making it less attractive.
He's certainly signed a contract, or he wouldn't be on the payroll! Not only that, but law enforcement officers are a huge legal special case with many of their rights curtailed. They are, for instance, LEOs 24hrs a day whether in uniform or not. It's not a 9-to-5 job with well defined boundaries. In addition, many states have statutory laws making intellectual property developed on the public dime public domain. Which this should be. All he can reasonably demand is overtime pay for the off-duty hours. But an employer can't be expected to lose the output of an employee just because the latter for whatever reason didn't file overtime.
If you're going to send messages, which is probably what you'll want at low bandwidths, there's got to be better and more efficient encodings and transmission protocols than CW. Off-hand, how about not sending the message in order so transmission errors don't result in consecutive symbols lost, and with CRC/ECC techniques and encapsulation to boost the chances of recovery (and reduce sensitivity to noise/loss). Isn't this the kind of thing that makes experimenting with 'moon bounce' and such fun in the first place? Experimenting with encodings, compressions, recovery methods, heuristics, homing algorithms, etc?
If I were part of the ham culture I'd be concerned that CW, by providing a predesigned but rather poor encoding and protocol standard, discourages innovation and entrenches mediocrity.
The info might not be the latest, but scroll down to the section on Energy Payback Time (EPBT): http://www1.eere.energy.gov/solar/pv_basics.html. Looks like a couple of years, give or take. If the technology with the 40% efficiency mentioned is equally cost-effective, then we're probably down under a year.
This just moves the burden of tamper-proofing from the machines in the booths to the one machine that scans the slips. If you can't tamper-proof the former, then you can't tamper-proof the latter either. Also, "not easily modifiable" isn't good enough, and is actually orthogonal to the issue; the question is not how hard it is to tamper with, but how hard it is to detect that it's been compromised. The equipment needs to be designed so that any compromise is obvious. Checkboxes on a printed card works quite nicely for this, it's perfectly clear whether a candidate has been removed from some ballots for instance. Any electronic machine needs to be equally obvious to the voter when it has been tampered with. I personally doubt this is even possible, and suspect it's the wrong approach.
Who wins? Polygraph vs chicken bones -- let the battle begin!
We trust ATMs because they are part of the banking system, and banks are trust corporations. Trust is their #1 product. If we didn't trust them we'd stash our money in our mattresses. People who don't trust banks do.
This is very different from a political system, where nobody trusts (and shouldn't trust) anybody else. The voter shouldn't trust the poll staff or equipment vendor, the poll staff shouldn't trust the voter or equipment vendor, etc. Basically, it's a system of mutual distrust. Such a system can only function on the principle of transparency and mutual verification. (Like Reagan said about the USSR, "trust but verify.") And this is where the electronic voting machine is incompatible: it's too complex and too concealed. There is no way for a voter to look and say yeah, the firmware looks trustworthy on this one. As opposed to punching a hole in a ballot, where it's pretty obvious whether the right hole got punched; the punch lever isn't going to jump and punch some other predetermined hole based on a statistical distribution, it's not complex enough to permit that degree of manipulation. In fact, the very act of bringing an unverifiable, complex machine into the center of the election process may be rooted in ignorance, but needs to be treated as a trojan horse. Unless it can be verified otherwise, an assumption needs to be made that it's compromised.
Elections is one place where 'simple and mechanical' is inherently better!
The problem isn't one of engineering, but of trust. Or, rather, mutual distrust. A system like that (e.g. election system) can't function in the long term without transparency and verification by all parties. This means the system needs to be a simple as possible, preferably with mechanical parts that can be eyeballed by anyone familiar with the system. After verification, the party (political parties, election officials, equipment maker) needs to be able to apply their seal to it. A firmware or electronic system in general is inherently unsuitable because of the difficulty of verification. It's also very difficult to adequately tamper proof, while DRM can be used it's not easy to check whether some other party has replaced the DRM system itself with a lookalike.
Another system built around mutual distrust was the cold war relationship between the USSR and the USA. Both parties accepted a high degree of spying and snooping, because it created some level of transparency and permitted mutual verification. Without rampant spying chances are good the paranoia had grown on both sides until it reached a point of war. 'Trust but verify'.
A lot of the cable/satellite HD isn't really HD (1080i) but lower resolution at as low as 4-6Mbps. Same bitrate as DVD, so don't be too surprised if it doesn't look much better. By HD they usually mean "a little better than DVD", but it's not a huge difference. Often it's worse that plain old DVD. It also varies from show to show, I've found the local cable HD sports broadcasts to be pretty good, but still not HD. I don't have a set yet, because I can't see the benefit either. I just looked around to see what's out there and wasn't impressed. The only way most of us are going to get HD content is through a player, so I'm simply sitting it out until there is one established format -- at which point 1080i sets should be more common, far less expensive, and demonstrably usable to their fullest with the players and STBs.
The software never sees content in plaintext, that would be utterly pointless.
A conditional access system needs to do all the backend work: interact with the subscriber and device databases, handle purchase requests, produce billing and audit data, look for fraud, revoke keys, assist with customer service, produce a gazillion different business reports, and of course issue keys. It needs to be tamper proof and self contained with a minimum of external interfaces. This is the space Sun has made an announcement in. They already had a framework to deliver the content plus the necessary client side support. On the client device that is requesting keys, managing keys, picking the right key out of a key block, and stuffing keys into or providing them to the secure decoder as needed. The delivery part is download or streaming.