This guy was complaining about having to buy 5 seats instead of 3 seats, and you're talking about licensing deals applicable to 5000 PCs?
There is no question that MS cost per PC can get low if you're a fortune-500 company. However, if you're a small dev shop with 5 PCs good luck getting that MS sales rep to return your calls.
For the average small business you're not going to get any kind of real support from MS without paying a chunk of cash. If you buy OSS from the main devs you are likely to be able to have a voice in product development even with a few seats.
I've dealt with both small vendors and large ones - the small ones are MUCH more responsive to user requests (if you're paying them well). The large ones often have a hard time getting their own act together. If you get in on the ground floor with a small vendor you might get a sweet licensing deal that gives you a great rate 10 years down the road while your competitors are just getting on board at steep rates. On the other hand, the large ones are more likely to be around in 10 years. Basically, if you deal with a small company you are more like an investor/owner and less like a customer. That means you share both the risk and the reward. Not something for everyone, but if you know what you're doing you can do well in this way.
Even then it depends - if the phone is in the employee's name it would probably be illegal unless the employee had signed a release. If the phone is in the company's name (especially if it is communal - such as an on-call phone) then it is likely to be legal. I doubt anybody would be dumb enough to use such a phone to leak info to the press, however...
Sometimes the cynic in me wants to say that society gives medals to firemen so that they'll keep on the job without asking for a pay hike... Why pay in cash when you can hold a ceremony every couple of years and put a medal on a grave?
Uh, wouldn't they need to show that the previous owner knew about the condition and failed to disclose it? If they didn't have health problems then most likely they didn't know about it, and usually that is all the defense you need for something like this. Granted, they can still file a nuiscence suit to try to get you to settle. Things like this are why most states have housing inspection clauses in their standard agreements...
If you're talking CO2 it is purely based on how much gas you use - 20gal is 10X 2 gal anyway you slice it.
If you're talking hydrocarbons then yes, on a per-gallon basis the 2-cycle is WAY more polluting. However, if you can get 10X the fuel efficiency it probably is cleaner overall, and there is no reason you can't slap a catalytic converter on it.
What ratio would you propose? 1000A at a mere 60KV (many industrial sites use 30KV distribution lines - strung out with 2' long insulators)? You'd need a monster cable and a 2' air gap around it. You probably can't go much over 1-2KV in practice (I'm no expert, but if 30KV towers have huge insulators to ensure an air gap you can't go anywhere near that), so now you're talking 10KA - that is a LOT of current. You'd need very low resistance to avoid melting the cable.
60 Megawatts is the kind of power that is transmitted over towers. There is no easy way to transmit that kind of power unless you have superconductors.
Agreed that you can trade-off volts for amps - but any way you slice it you have a big problem at those power levels.
Ten guys sleeping on benches all day - public nuisence.
15 guys in a race to build 4-bedroom colonial homes on the park - tell me again what the park was there for?
Pirate radio is not a problem only because it is rare. Once there are 50 in every town there will be a mess. Sure, that won't happen overnight, but if you give public blessing to this use of the airwaves it WILL happen (especially when Walmart starts selling FM transmitters for $200).
You missed another aspect of the used market which impacts the first-sale market - value. A game which can be resold is more valuable than a game that cannot. Sure, people don't think about it much now, but that is only because all the titles can be resold. If you end up with a mix people will realize that a resellable game is cheaper since once you get tired of it you can recoup some of your initial investment by selling it. A game which cannot be resold just sits on the shelf unused.
What is the 5-year cost of owning a new car? It isn't just the sticker price plus the cost of gas/maintenance/etc. It is all that MINUS the value on sale 5 years later - and that number is a BIG part of the equation as it makes the replacement car a whole lot cheaper to purchase.
Console publishers who don't allow resale will find consumers willing to pay less for their games. Now, it may well be that they more than compensate for this by getting rid of the resale market, but I think that this will only work for wildly popular games. Initially consumers will pay big bucks not realizing the hidden cost of this scheme, but if it takes off consumers will figure it out...
Perhaps they want to, and by threatening studios they hope to get them to allow them to distribute media in this way.
The difference between DVDs/music and photos is that the former are copyrighted (well so are the latter, but by the customer so if the customer wants to download them there is no issue). Walmart can't just choose to distribute movies over the net and then cut the studios a check. And they couldn't compete with Apple if they had to buy a boxed edition for every one they sell online (and even then courts might not like them unilaterally changing the format - even if it results in NO lost DVD sales).
The problem is that copyright gives the studios a lot of control over HOW their movies are sold - and not just that they get the profits when they are sold.
But no one seems to recognize that HP was legally obligated to find and address the source of the leak. Leaks of financial info governed by the SEC are illegal, and could have landed HP in far larger hot water than this. (Think insider trading, etc)
There is a simple mechanism for doing this - file a notice with the SEC indicating that they suspect info is being disclosed in a manner not authorized by the company, and request an investigation. The SEC has the power of government to request warrants and perform the investigation in a perfectly legal and ethical manner.
Or, if they felt it were appropriate they could file lawsuits against anybody they had cause to suspect, and issue subpoenas and discovery requests for phone records, computers, etc. This would also be legal and ethical - as long as the judge applied scrutiny to the request and gave defendants an opportunity to defend themselves (ie no filing of lawsuits just to find out who disclosed things, at which point the suit is dropped and SLAPPs or blacklisting starts).
The point is that there is a right way and wrong way to carry out an investigation. The difference between the two is DUE PROCESS. The folks who get to decide what due process is are the people we elect into positions to exercise this authority - not the board of HP or some private investigator.
And the press also had to know leaks of this nature were illegal. Where is their accountability?
They can be sued, or the SEC can investigate if warranted.
However, I don't see how a press story would be a violation of SEC rules. SEC rules prevent people from ACTING on insider information. They don't prevent people from publishing it. Once it is published it is no longer insider information anyway. The spirit of the law is to give everybody equal access to information. As long as the reporter publishes it in the open press and not in some newletter with a circulation of 50 there shouldn't be any problems.
Under what circumstances would it be legal at all?
I'd say that it would be legal under only a few sets of circumstances:
1. The owner of the phone signed a release allowing the phone company to release the information. 2. A court signed a warrant or subpoena for the necessary information, which was presented to the phone company. Said document would require some sort of probable cause and a limit of scope.
I can't think of any others. Just about anything else would probably be unethical at best - and almost certainly illegal. If the company had either #1/2 they would have come out with it by now to avoid getting dragged through the mud. Somebody violated the law...
Well, in any case GPLv3 does not prevent voting machines running GPLv3 code from having hardware that verifies them against keys that are not publicly disclosed. You only need to disclose the key if you distribute the software. If you distribute the software so that it can run on any hardware, and then on one specific piece of hardware you require signed binaries, but you don't distribute this hardware to the public, then you shouldn't have to distribute the keys to the public, because the keys aren't required for the intended use of the software to the people it is being distributed to. Ie, you essentially have two different products - one is a software-only product that you're distributing that runs anywhere (no keys needed), and one is a software/hardware hybrid that you're only distributing to state governments, and those states get a copy of the necessary keys. Each product includes all the keys necessary to run it.
Only the owner of the machines need receive the keys (which is fine - that would be the state board of elections most likely). The machines aren't being sold to the public, so the public doesn't need to have the ability to modify the machines.
while there is a possibility for recount, it is difficult and expensive to call for one
Not really. Just pick some district whose results seem skewed from the exit poll data, and do a recount on a few wards. If they all count up right, then there is probably nothing going wrong. If there is a major discrepancy it won't be hard to get a recount ordered for the entire state, and a major criminal investigation started. In fact, the use of paper audit trails should always be accompanied by an automatic partial recount (possibly even a 100% recount - announce the winner on election night but certify the counts a month later).
Redhat can't distribute the kernel without the source, under the GPLv3 this would include the key to sign the binary.
This is incorrect, unless the redhat kernel does a self-test to see if it is signed. If Redhat Linux does not in any way use the key, then they don't need to include it. Now, if Redhat somehow verified the integrity of itself then it would be required to use the key.
The GPL v3 simply says that if you somehow restrict a user to running only signed binaries, then you must provide the signing key. Redhat does none of this, and therefore they don't need to provide their keys. Tivo does, and therefore Tivo must supply the appropriate key.
Here is the wording (emphasis mine):
The Corresponding Source also includes any encryption or authorization keys necessary to install and/or execute modified versions from source code in the recommended or principal context of use, such that they can implement all the same functionality in the same range of circumstances.
Redhat's keys are not necessary to run Redhat's software in the recommended or principal context of use. On the other hand, if Tivo redistributes the software and requires the files be signed, then the keys ARE required to run the software in the recommended/principal context of use. Since the intended use of the two distributions are different, the requirement to package keys is different.
I certainly think that trusted computing and DRM have places where they are important - medical, financial, and security environments come to mind immediately. The GPLv2 allows Linux and GNU software to be constructed & run in these environments.
How is that? If I sell somebody a DRMed heart monitor that does a self-integrity test, I must also give them the ability to defeat that test. That doesn't mean that they have to pass that ability on to anybody else, unless they give them the monitor. So, the hospital has the ability to sabotage its own equipment - that doesn't matter since they're not interested in doing this (otherwise why would they spend money on equipment that does self-integrity tests), and they don't have to pass this on to anyone else (their workers, patients, etc). It is the owner of the software that gets the keys to the kingdom in GPLv3 - it doesn't mean that banks need to give access codes to their ATMs out to anybody who has an ATM card.
The GPL only applies when you distribute software. If you don't distribute it, you don't need to pass on encryption keys/etc. If somebody steals it from you then you also don't need to pass on these keys. So, DRM can be used by the owner of a system to keep that system secure - which is a use of technology that I doubt the FSF would take issue with. What it prevents is the manufacturer of the system from keeping it locked up from being changed by the owner.
For me it's interesting to note that RMS has said that there would have been no issues with TIVO if they had used ROM instead of FLASH to store the software.
I don't see why this should be the case - software is software no matter what media it is stored on. The kernel is still copyrighted, so a license applies to its distribution. There is no reason other than the difficulty factor that a Tivo owner couldn't swap out some ROM chips - just look at console modders...
DRM is inherently a poor way to secure a voting machine.
If I'm a voter - how do I know the DRM securing my voting machine is intact? How do I know that instead of just reflashing the software on the machine that somebody didn't instead replace the DRM-chip and reflash the software? How do I know the folks who loaded the software in the first place are legit?
The most straightforward solution to e-voting confidence is to let the voter SEE the votes. WITH HIS EYES! Print them out on paper, and either use the paper copy as the official tally, or at least keep it around for auditing (and audit some percentage of the votes - possibly 100%). Any idiot can understand how it works, and the software keeps invalid ballots from being generated (ie votes for two people for president and a bunch of judges trying to figure out "voter intent"). The worst somebody could do by tampering with the software is to generate invalid ballots, but those will get spotted very quickly by the voters.
The goal should be to make things open and accessible - not closed and hidden. DRM does not make things open to the user.
If the environment (ie Tivo) will only execute a signed binary the key must be released to allow someone to run the modified binary Many think this is an unacceptable consequence.
Yes, but the obligation to release the signing key is on Tivo - not some kernel developer. If Tivo can't obtain the key they want to use, then they can't distribute the software at all. Therefore, Tivo won't be able to verify signatures generated by 3rd-parties - only their own signatures. This really doesn't diminish data security - only a loophole that would otherwise let them employ DRM-like techniques while claiming innocence.
Nothing that somebody else does can force you to do anything. If my next-door neighbor offers my house to somebody and they shake on the deal, it means nothing to me. Likewise, if Tivo promises in a GPLv3 license that they'll turn over Linus's PGP key, that does not in any way force Linus to give it up - instead Tivo will end up violating the GPL when they can't come up with it themselves.
Uh, it seems unintuitive to me that a little needle can possibly vibrate at 25+ kHz without tremendous attenuation. Analog or not bandwidth is still bandwidth (the term was actually invented long before ditigal data transmission). The bandwidth of a vibrating needle is not going to be that high - a simple consequence of inertia.
Now, I'll admit that this is just intuition on my part, and perhaps somebody will show some math to prove that I'm wrong (or a real-world measurement). I'd be very surprised if you can cut a groove into a record and play it back and get a frequency of 100kHz to show up at all.
As far as frequencies of 100kHz having an impact on the audible range - it could only happen if something resonates at high frequency and then retransmits the sound at a lower frequency - perhaps an object in the listening room. For example, you can't hear 5 Hz, but you can hear 1 kHz bangs given off at 5 Hz modulation as some metal object vibrates. You might also have some impact if you get beats from the interference of high-frequency sounds. However, I'd question how well those frequencies can actually be generated by needles/speakers/etc.
Again, I'm not an audio engineer, but I'd really need to see the physics to be convinced that these effects are real - sound is a wave like any other, and is subject to the rules of both physics (inertia, dampening, resonance, etc) and information theory (you can take waves of two frequencies, add them together, and split them apart again).
Hey, if I'm trying to sell something I might bend over backwards for IE - that's just realpolitik.
But, that isn't how it is supposed to work.
If intel decided to mess up their opcodes on each generation of their CPUs the solution wouldn't be to build binaries that contain 47 complete copies of the software.
The whole reason you have standards is to avoid having to build every package on every vendor's box...
I agree with you on that point - you should code to the standard and making it look right is the browser's problem.
However, I do want to comment that the threshold box tends to load outside of the window on konqueror - which is ACID2 compliant in the version I am using. If I hit the top link to reload just the comments it works fine.
I tend to doubt that 8-way dual opterons would be enough, although I can't say I'm an expert in this.
I'm guessing that the necessary hardware would be a combination of DSPs, CPUs, lots of RAM, etc. They actually exist and are commercially available - how do you think that live HD programming gets compressed in the first place? They're just WAY too expensive for consumer use.
Hey - I don't agree with data-mining phone records without a warrent. In fact, I'd suggest that this is actually a weaker position than capital punishment as it does not even have a basis in law. However, I'd still argue that the NSA tapping phones is distinctly different from an HP investigative team doing the same, as the NSA is in fact a government agency, and it is under orders from the president, which while not appropriate at least has some kind of democratic representation behind it.
And if you're really opposed to US government intrusion in life there are candidates on the ballot who would take your position - most are 3rd parties (and don't get me started on the problems with the plurality-based voting system). Even with something like proportional democracy it seems like in Europe still has trouble with this issue - so it isn't as simple as election reform.
It is all about the interface. Try hooking your PC up to a TV without a monitor, and with only a remote control for input. Good luck with firefox...
The whole idea behind myth is that it is optimized to operate like a VCR/DVD/Tivo/etc. Sure, half the modules aren't needed if you're just running it on a computer with a high-res monitor, keyboard, and mouse. In fact, mythtv can be annoying on such a device since you can't just use the mouse to select stuff (mythweb instead is a good interface for programming from a PC)...
Slashdot Mirror
This guy was complaining about having to buy 5 seats instead of 3 seats, and you're talking about licensing deals applicable to 5000 PCs?
There is no question that MS cost per PC can get low if you're a fortune-500 company. However, if you're a small dev shop with 5 PCs good luck getting that MS sales rep to return your calls.
For the average small business you're not going to get any kind of real support from MS without paying a chunk of cash. If you buy OSS from the main devs you are likely to be able to have a voice in product development even with a few seats.
I've dealt with both small vendors and large ones - the small ones are MUCH more responsive to user requests (if you're paying them well). The large ones often have a hard time getting their own act together. If you get in on the ground floor with a small vendor you might get a sweet licensing deal that gives you a great rate 10 years down the road while your competitors are just getting on board at steep rates. On the other hand, the large ones are more likely to be around in 10 years. Basically, if you deal with a small company you are more like an investor/owner and less like a customer. That means you share both the risk and the reward. Not something for everyone, but if you know what you're doing you can do well in this way.
Even then it depends - if the phone is in the employee's name it would probably be illegal unless the employee had signed a release. If the phone is in the company's name (especially if it is communal - such as an on-call phone) then it is likely to be legal. I doubt anybody would be dumb enough to use such a phone to leak info to the press, however...
Sometimes the cynic in me wants to say that society gives medals to firemen so that they'll keep on the job without asking for a pay hike... Why pay in cash when you can hold a ceremony every couple of years and put a medal on a grave?
Ok, how can a slashdot definition be complete without the obvious:
4. Command used to send a message to all available operators. (As in Write ALL OPeratorS).
Ok, so they don't know how to spell and dropped the s...
Uh, wouldn't they need to show that the previous owner knew about the condition and failed to disclose it? If they didn't have health problems then most likely they didn't know about it, and usually that is all the defense you need for something like this. Granted, they can still file a nuiscence suit to try to get you to settle. Things like this are why most states have housing inspection clauses in their standard agreements...
Well, depends on the type of pollution.
If you're talking CO2 it is purely based on how much gas you use - 20gal is 10X 2 gal anyway you slice it.
If you're talking hydrocarbons then yes, on a per-gallon basis the 2-cycle is WAY more polluting. However, if you can get 10X the fuel efficiency it probably is cleaner overall, and there is no reason you can't slap a catalytic converter on it.
What ratio would you propose? 1000A at a mere 60KV (many industrial sites use 30KV distribution lines - strung out with 2' long insulators)? You'd need a monster cable and a 2' air gap around it. You probably can't go much over 1-2KV in practice (I'm no expert, but if 30KV towers have huge insulators to ensure an air gap you can't go anywhere near that), so now you're talking 10KA - that is a LOT of current. You'd need very low resistance to avoid melting the cable.
60 Megawatts is the kind of power that is transmitted over towers. There is no easy way to transmit that kind of power unless you have superconductors.
Agreed that you can trade-off volts for amps - but any way you slice it you have a big problem at those power levels.
Squatters are a good example.
One guy who sleeps on a park bench - no big deal.
Ten guys sleeping on benches all day - public nuisence.
15 guys in a race to build 4-bedroom colonial homes on the park - tell me again what the park was there for?
Pirate radio is not a problem only because it is rare. Once there are 50 in every town there will be a mess. Sure, that won't happen overnight, but if you give public blessing to this use of the airwaves it WILL happen (especially when Walmart starts selling FM transmitters for $200).
You missed another aspect of the used market which impacts the first-sale market - value. A game which can be resold is more valuable than a game that cannot. Sure, people don't think about it much now, but that is only because all the titles can be resold. If you end up with a mix people will realize that a resellable game is cheaper since once you get tired of it you can recoup some of your initial investment by selling it. A game which cannot be resold just sits on the shelf unused.
What is the 5-year cost of owning a new car? It isn't just the sticker price plus the cost of gas/maintenance/etc. It is all that MINUS the value on sale 5 years later - and that number is a BIG part of the equation as it makes the replacement car a whole lot cheaper to purchase.
Console publishers who don't allow resale will find consumers willing to pay less for their games. Now, it may well be that they more than compensate for this by getting rid of the resale market, but I think that this will only work for wildly popular games. Initially consumers will pay big bucks not realizing the hidden cost of this scheme, but if it takes off consumers will figure it out...
Perhaps they want to, and by threatening studios they hope to get them to allow them to distribute media in this way.
The difference between DVDs/music and photos is that the former are copyrighted (well so are the latter, but by the customer so if the customer wants to download them there is no issue). Walmart can't just choose to distribute movies over the net and then cut the studios a check. And they couldn't compete with Apple if they had to buy a boxed edition for every one they sell online (and even then courts might not like them unilaterally changing the format - even if it results in NO lost DVD sales).
The problem is that copyright gives the studios a lot of control over HOW their movies are sold - and not just that they get the profits when they are sold.
But no one seems to recognize that HP was legally obligated to find and address the source of the leak. Leaks of financial info governed by the SEC are illegal, and could have landed HP in far larger hot water than this. (Think insider trading, etc)
There is a simple mechanism for doing this - file a notice with the SEC indicating that they suspect info is being disclosed in a manner not authorized by the company, and request an investigation. The SEC has the power of government to request warrants and perform the investigation in a perfectly legal and ethical manner.
Or, if they felt it were appropriate they could file lawsuits against anybody they had cause to suspect, and issue subpoenas and discovery requests for phone records, computers, etc. This would also be legal and ethical - as long as the judge applied scrutiny to the request and gave defendants an opportunity to defend themselves (ie no filing of lawsuits just to find out who disclosed things, at which point the suit is dropped and SLAPPs or blacklisting starts).
The point is that there is a right way and wrong way to carry out an investigation. The difference between the two is DUE PROCESS. The folks who get to decide what due process is are the people we elect into positions to exercise this authority - not the board of HP or some private investigator.
And the press also had to know leaks of this nature were illegal. Where is their accountability?
They can be sued, or the SEC can investigate if warranted.
However, I don't see how a press story would be a violation of SEC rules. SEC rules prevent people from ACTING on insider information. They don't prevent people from publishing it. Once it is published it is no longer insider information anyway. The spirit of the law is to give everybody equal access to information. As long as the reporter publishes it in the open press and not in some newletter with a circulation of 50 there shouldn't be any problems.
Under what circumstances would it be legal at all?
I'd say that it would be legal under only a few sets of circumstances:
1. The owner of the phone signed a release allowing the phone company to release the information.
2. A court signed a warrant or subpoena for the necessary information, which was presented to the phone company. Said document would require some sort of probable cause and a limit of scope.
I can't think of any others. Just about anything else would probably be unethical at best - and almost certainly illegal. If the company had either #1/2 they would have come out with it by now to avoid getting dragged through the mud. Somebody violated the law...
Well, in any case GPLv3 does not prevent voting machines running GPLv3 code from having hardware that verifies them against keys that are not publicly disclosed. You only need to disclose the key if you distribute the software. If you distribute the software so that it can run on any hardware, and then on one specific piece of hardware you require signed binaries, but you don't distribute this hardware to the public, then you shouldn't have to distribute the keys to the public, because the keys aren't required for the intended use of the software to the people it is being distributed to. Ie, you essentially have two different products - one is a software-only product that you're distributing that runs anywhere (no keys needed), and one is a software/hardware hybrid that you're only distributing to state governments, and those states get a copy of the necessary keys. Each product includes all the keys necessary to run it.
Only the owner of the machines need receive the keys (which is fine - that would be the state board of elections most likely). The machines aren't being sold to the public, so the public doesn't need to have the ability to modify the machines.
while there is a possibility for recount, it is difficult and expensive to call for one
Not really. Just pick some district whose results seem skewed from the exit poll data, and do a recount on a few wards. If they all count up right, then there is probably nothing going wrong. If there is a major discrepancy it won't be hard to get a recount ordered for the entire state, and a major criminal investigation started. In fact, the use of paper audit trails should always be accompanied by an automatic partial recount (possibly even a 100% recount - announce the winner on election night but certify the counts a month later).
This is incorrect, unless the redhat kernel does a self-test to see if it is signed. If Redhat Linux does not in any way use the key, then they don't need to include it. Now, if Redhat somehow verified the integrity of itself then it would be required to use the key.
The GPL v3 simply says that if you somehow restrict a user to running only signed binaries, then you must provide the signing key. Redhat does none of this, and therefore they don't need to provide their keys. Tivo does, and therefore Tivo must supply the appropriate key.
Here is the wording (emphasis mine):
Redhat's keys are not necessary to run Redhat's software in the recommended or principal context of use. On the other hand, if Tivo redistributes the software and requires the files be signed, then the keys ARE required to run the software in the recommended/principal context of use. Since the intended use of the two distributions are different, the requirement to package keys is different.
I certainly think that trusted computing and DRM have places where they are important - medical, financial, and security environments come to mind immediately. The GPLv2 allows Linux and GNU software to be constructed & run in these environments.
How is that? If I sell somebody a DRMed heart monitor that does a self-integrity test, I must also give them the ability to defeat that test. That doesn't mean that they have to pass that ability on to anybody else, unless they give them the monitor. So, the hospital has the ability to sabotage its own equipment - that doesn't matter since they're not interested in doing this (otherwise why would they spend money on equipment that does self-integrity tests), and they don't have to pass this on to anyone else (their workers, patients, etc). It is the owner of the software that gets the keys to the kingdom in GPLv3 - it doesn't mean that banks need to give access codes to their ATMs out to anybody who has an ATM card.
The GPL only applies when you distribute software. If you don't distribute it, you don't need to pass on encryption keys/etc. If somebody steals it from you then you also don't need to pass on these keys. So, DRM can be used by the owner of a system to keep that system secure - which is a use of technology that I doubt the FSF would take issue with. What it prevents is the manufacturer of the system from keeping it locked up from being changed by the owner.
For me it's interesting to note that RMS has said that there would have been no issues with TIVO if they had used ROM instead of FLASH to store the software.
I don't see why this should be the case - software is software no matter what media it is stored on. The kernel is still copyrighted, so a license applies to its distribution. There is no reason other than the difficulty factor that a Tivo owner couldn't swap out some ROM chips - just look at console modders...
DRM is inherently a poor way to secure a voting machine.
If I'm a voter - how do I know the DRM securing my voting machine is intact? How do I know that instead of just reflashing the software on the machine that somebody didn't instead replace the DRM-chip and reflash the software? How do I know the folks who loaded the software in the first place are legit?
The most straightforward solution to e-voting confidence is to let the voter SEE the votes. WITH HIS EYES! Print them out on paper, and either use the paper copy as the official tally, or at least keep it around for auditing (and audit some percentage of the votes - possibly 100%). Any idiot can understand how it works, and the software keeps invalid ballots from being generated (ie votes for two people for president and a bunch of judges trying to figure out "voter intent"). The worst somebody could do by tampering with the software is to generate invalid ballots, but those will get spotted very quickly by the voters.
The goal should be to make things open and accessible - not closed and hidden. DRM does not make things open to the user.
If the environment (ie Tivo) will only execute a signed binary the key must be released to allow someone to run the modified binary
Many think this is an unacceptable consequence.
Yes, but the obligation to release the signing key is on Tivo - not some kernel developer. If Tivo can't obtain the key they want to use, then they can't distribute the software at all. Therefore, Tivo won't be able to verify signatures generated by 3rd-parties - only their own signatures. This really doesn't diminish data security - only a loophole that would otherwise let them employ DRM-like techniques while claiming innocence.
Nothing that somebody else does can force you to do anything. If my next-door neighbor offers my house to somebody and they shake on the deal, it means nothing to me. Likewise, if Tivo promises in a GPLv3 license that they'll turn over Linus's PGP key, that does not in any way force Linus to give it up - instead Tivo will end up violating the GPL when they can't come up with it themselves.
Uh, it seems unintuitive to me that a little needle can possibly vibrate at 25+ kHz without tremendous attenuation. Analog or not bandwidth is still bandwidth (the term was actually invented long before ditigal data transmission). The bandwidth of a vibrating needle is not going to be that high - a simple consequence of inertia.
Now, I'll admit that this is just intuition on my part, and perhaps somebody will show some math to prove that I'm wrong (or a real-world measurement). I'd be very surprised if you can cut a groove into a record and play it back and get a frequency of 100kHz to show up at all.
As far as frequencies of 100kHz having an impact on the audible range - it could only happen if something resonates at high frequency and then retransmits the sound at a lower frequency - perhaps an object in the listening room. For example, you can't hear 5 Hz, but you can hear 1 kHz bangs given off at 5 Hz modulation as some metal object vibrates. You might also have some impact if you get beats from the interference of high-frequency sounds. However, I'd question how well those frequencies can actually be generated by needles/speakers/etc.
Again, I'm not an audio engineer, but I'd really need to see the physics to be convinced that these effects are real - sound is a wave like any other, and is subject to the rules of both physics (inertia, dampening, resonance, etc) and information theory (you can take waves of two frequencies, add them together, and split them apart again).
Hey, if I'm trying to sell something I might bend over backwards for IE - that's just realpolitik.
But, that isn't how it is supposed to work.
If intel decided to mess up their opcodes on each generation of their CPUs the solution wouldn't be to build binaries that contain 47 complete copies of the software.
The whole reason you have standards is to avoid having to build every package on every vendor's box...
Uh, that would be the purpose of the else statement, wouldn't it?
I agree with you on that point - you should code to the standard and making it look right is the browser's problem.
However, I do want to comment that the threshold box tends to load outside of the window on konqueror - which is ACID2 compliant in the version I am using. If I hit the top link to reload just the comments it works fine.
So this isn't an IE-only issue...
I tend to doubt that 8-way dual opterons would be enough, although I can't say I'm an expert in this.
I'm guessing that the necessary hardware would be a combination of DSPs, CPUs, lots of RAM, etc. They actually exist and are commercially available - how do you think that live HD programming gets compressed in the first place? They're just WAY too expensive for consumer use.
Hey - I don't agree with data-mining phone records without a warrent. In fact, I'd suggest that this is actually a weaker position than capital punishment as it does not even have a basis in law. However, I'd still argue that the NSA tapping phones is distinctly different from an HP investigative team doing the same, as the NSA is in fact a government agency, and it is under orders from the president, which while not appropriate at least has some kind of democratic representation behind it.
And if you're really opposed to US government intrusion in life there are candidates on the ballot who would take your position - most are 3rd parties (and don't get me started on the problems with the plurality-based voting system). Even with something like proportional democracy it seems like in Europe still has trouble with this issue - so it isn't as simple as election reform.
It is all about the interface. Try hooking your PC up to a TV without a monitor, and with only a remote control for input. Good luck with firefox...
The whole idea behind myth is that it is optimized to operate like a VCR/DVD/Tivo/etc. Sure, half the modules aren't needed if you're just running it on a computer with a high-res monitor, keyboard, and mouse. In fact, mythtv can be annoying on such a device since you can't just use the mouse to select stuff (mythweb instead is a good interface for programming from a PC)...