Somebody ought to label the initial "What about bans?" irrelevant and label its poster a troll. I can't do it, but this one is not only unrelated to the questions of civil liberties raised, but arguably the product of either stupidity or payola.
Humorous reminders need not be malicious
on
Darwin Awards 2006
·
· Score: 1, Insightful
I have a very good friend who, despite knowing all the reasons why she shouldn't, smokes. This years "Stubbed Out" made an impression. If the effect is to get one person to stop smoking (or doing other really stupid things like igniting gasoline in florescent tubes), they server a valuable social function.
Obviously the Tesla effect is not what's interesting here. Its the direction of power to the device that is interesting. You have this big power sheet and it only sends power where there is a device that needs it.
The obvious place for this power supply is your chair and bed. You can sit down with your cell phone in your pocket and have it recharge while you work at your computer.
The big deal application of this kind of technology is probably computer implants. Its hard to change the batteries of a computer that's embedded in your thigh, but now it can recharge while you sleep.
I see a big market for upholstery and bedding growing out of this device.... maybe...
So I'll take my Egg McMuffin with the Florescent Green Canadian Bacon. Hmmm. I think not.
Here's an accomplishment that will create a real conundrum for the Green Party. Is the pig a symbol to be cherished, a genetic freak to be feared, or some combination.
As long as the interfaces are standard and the software can be ported from one machine to the next, there shouldn't be a problem. The latter is, of course, an argument against using Windows as the platform, but there are good Windows emulators on Linux, so all is not lost.
Not at all. There is no necessary fault in being Republican or Democrat or in the unrelated categories of being liberal or conservative. Republicans do win, and should win, elections without resorting to chicanery. Democrats do, and should do, the same.
Where, however, substantial doubt can be cast on an election outcome, due diligence should be used to resolve the issue to the extent that it is possible to do so. If there was a problem with the voting machines in this district that caused a huge number of votes to evaporate, it should be investigated. If the problem can be documented, corrective actions should be taken. That's all that this litigation is asking for.
The problem here isn't a Republican or Democratic problem. It is a problem of ensuring the election process is conducted correctly.
The problem with this judges decision is that it (in my view incorrectly) stops the due diligence process for exactly the wrong reasons. A corporations proprietary interests should NEVER trump the electoral process or throw its results into doubt.
Which is exactly why all election software should be publicly verifiable open source.
Good points, but resolvable. I believe, at this point, that the only means by which we can provide full confidence in the system is for people to be able to confirm their votes after the fact. There may be no way of preventing vote buying in such a system, but I believe that the take away can be sufficiently secured as to make intimidation difficult and vote buying more readily detectable. In "There ought to be a low", below, I describe the copy as encrypted. Assuming the copy is made with a public key, the voter should be able to check their vote while still in the booth and then later (after the votes have been counted) online. They would be equally able to present the copy to a courtroom if needed, but the copy would not be human readable, immediately accessible outside of the voting booth, or accessible at any point without the keys.
Such a system would not stop vote buying, which happens even without a means of verification. Most people take their votes more seriously than that, but the confirmation system would allow the identification of any machine that verifies a large number of ballots, which ought to make it easier to detect and punish vote buying. There are no technical panaceas to resolve ill behavior, but it ought to be possible for the 18,000 people who may have been disenfranchised in the election documented here to check to make sure their votes were properly counted.
It is generally much easier for us to critique new systems, like digital systems are today, than it is to critique well established systems, which often have glaring flaws of their own. We don't see the flaws in older systems because we have internalized and/or have been socialized into what is good about them and have accepted the glaring flaws of existing systems as normal, unavoidable, or offset by the positives.
There are many kinds of data insecurity. Recent demonstrations of the inherent "insecurity" of digital information focus on particular kinds of insecurity, including:
the possibility that data can be undetectably replaced (the reason why voting machines should be secure and off network).
the possibility that data can be undetectably miswritten (the reason why reviewing electoral source code matters).
Punch card systems (which are pseudo-digital) have been demonstrated to be open to other kinds of insecurity:
verifiability by the voter (a problem in South Florida in 2000)
difficult to interpret and uninterpretable votes (also a problem in South Florida in 2000 - hanging chads)
disappearance of cards (Jacksonville, Fl in 2000)
Paper votes have historically been open to still other kinds of insecurity
extra ballots stuffed into the box which cannot be shown not to be legitimate votes
balloting by the voting dead (a particular problem in Chicago many years ago)
registration in multiple districts (e.g. the reality of vote early and often)
uninterpretable or disputed marks on the ballots
disappearing ballots
the possibility that someone can identify your ballot from your handwriting (hence abnegating the value of a secret ballot) without your assent.
I would argue that the mechanical machines we've been using in New York for generations resolve most of the forms of insecurity outlined above. They certainly provide a publicly verifiable ballot and a reasonably good (two level) audit trail, but they aren't perfect either, as there is no way for an individual voter to confirm that their vote was recorded correctly.
A properly done digital voting system can provide much better security than any existing system has managed. That won't happen, however, if we let the scandalously operated proprietary voting machines of the present stop us from looking for better ways to do things.
Paper ballots can dissapear into the trunk of a car and never be seen again, as reportedly happened in Jacksonville, FL in the 2000 election.
You are right, simple is better, but wouldn't it have been better if the observers for both parties could have taken a copy of the voting machine with them on a USB dongle at the end of the day? Wouldn't it be better if voters could carry a copy of their vote home with them so that it could be accounted for after the fact if necessary?
In this case, simple would be simple open source code that interpreted a simple human readable (e.g. XML) ballot specification with votes verifiable at several levels.
Pencil and paper simply provide other means of stuffing the ballot box, ways that are much harder to detect.
A good hacker who can engineer access can always get access to the software, whether open source or proprietary. Obscurity is no defense except in the courts, where hacking the code would be illegal and therefore inadmissible (and probably worse).
The big lesson of open source is that lots of eyes tend to make software more secure because concerned reviewers of the code find the flaws and, with the agreement of the large community, fix them. The security of Linux is far better than that of windows and, judging from comments about Vista, far less intrusive.
Your comment directly contradicts the article that starts the discussion.
This is a case where a Democrat is alleging voting irregularities and in which the Democratic party is allowing the Republican to take a seat in Congress despite the fact that the outcome of the election is still in the courts.
Oh, and by the way, Clinton took several shots at taking out Bin Laden (cruise missile attacks on Al Qaeda training camps in Afghanistan, as I recall). They didn't work, but they were well documented in the newspapers along with Republican assertions that Clinton was "wagging the dog".
This is an idea that won't die, and the car companies (Ford, GM, Daimler Chrysler, Toyota, Honda, etc) have a long history of talking to the big computer companies (IBM and now Microsoft) when they explore it. If Ford went to Gates, its because he regards Gates as a premier expert. If he focused on Windows, its simply because its widely used and has a logo that will go nicely on the windows of your car. Hobbyists have been tricking out their cars with computers for years, and the DVD players that are hanging in the car of every family who takes a trip with a five year old are already turning into Video Game stations.
The good news is that, when cars start arriving in volume that have computers built in, there will quickly be:
Fifty ports of Linux, almost all of which really will put the car a generation ahead.
A quickly evolving industry for aftermarket computer upgrades, including replacement. This industry will be particularly important when cars get to being two years or so old, because the cars functional life will undoubtedly exceed that of the computer. That will make such upgrades particularly important to dealers of used cars.
CPU and memory comparisons added to blue book values.
None of this is intended to discount the very real problems associated with actually using such a computer while you are driving, but if something is going to happen eventually, better to seriously consider its implications, if only because there may be profit in doing so.
The software for electronic voting is not complicated. I've written software in the general category (survey software) myself. Good survey software should have only two parts:
The program that displays the ballot and records the votes.
A specification of the ballot
The first part should be open source. There is no reason for it to be proprietary except to hide its workings or pretend that its complicated (which it isn't). It should be sufficiently stable that it can run as a ROM. If it isn't, it shouldn't be in use. If it isn't simple, moreover, than more attention has been paid to bells and whistles than to the function of the software, which should:
tally the results, as they occur, in a simple file system structure. A database is overkill for this purpose. An appended spreadsheet file would be more sensible.
save a copy of each ballot cast as a discrete file with a unique identifier
print an copy of the ballot cast for retention, including the unique identifier
print an encrypted copy of the ballot cast that the voter can take with them, including the unique identifier.
The result is a tally sheet that can be immediately checked at the end of voting and three levels by which the result can be double checked> The third level (a copy of the ballot cast that the voter takes with them) will allow individuals to ensure that their ballot was counted correctly.
All of which is a run up to saying that there ought to be a law that requires voting software to be open source such that every level of the process is transparent to voters.
Obviously someone who is unaware of the millions of machines that are routinely overheated by overload... most machines running graphics intensive applications and all machines running BOINC.
Bad thinking, but wishful thinking often is.
Davis
One of the problems associated with giving a good technical reply is that we don't really have much of a technical nature to work with here. All we really have, unless we resort to secondary materials (one useful source is the Wikipedia article on MediaSentry: URL:http://en.wikipedia.org/wiki/MediaSentry), is a Curriculum Vita and a legally filed declaration by Dr. Doug Jacobson.
I don't see any glaring faults in the Curriculum Vita. Doug may have some biases relative to P2P file sharing (invalid biases in my view: the first mark of a successful new medium of communication is the assertion that it is being used for pornography), but those biases hardly seem relevant to this case. His background in networking and security is at least as good as some 10,000 other people who might be called as expert witnesses, and he's willing to testify on behalf of the RIAA, which very few of those 10,000 people would be willing to do, mostly based on the well justified belief that the RIAA is using these suits to defend oligarchical control of music and video rather than defend intellectual property rights. He certainly has a background in security, networking, and computer forensics that would make him a credible witness from a technical perspective. The only fault I might point at is that it demonstrates no specific knowledge or evaluation of MediaSentry. His background and patent activity in the area of detecting and blocking P2P streams should, however, make him reasonably qualified to evaluate MediaSentry. By the way, it is probably worth asking him to evaluate MediaSentry. His evaluation may not glowing.
Dr. Jacobson's declaration is another matter entirely. It uses nine pieces of evidence, none of which are available to us here, to comment on three points in the connectivity chain that purportedly ties a computer to the Kazaa file sharing system, allegedly for the purposes of both downloading and sharing music files. Those three points are (a) the computer itself, as represented by its disk drive image, (b) the computers connection to Verizon Internet Services and, by implication, to the Internet, and (c) the detection and documentation of Internet activity associated with a specific set of IP addresses by MediaSentry. That is a major collapsing of the points of connectivity between a computer and a remote server. If I were investigating (and this is an area where I have patents) the connection points from a computer to a remote server, I'd identify quite a few more, starting with the intermediate modem, cable modem, or DSL modem and the downstream DHCP server maintained by Verizon, but this is all we have. I will return to this issue below, but there is value in addressing each of the three points in the chain that Dr. Jacobson does identify, and I'll return to this as we discuss the second issue, the computer's connection to Verizon.
First, the hard drive. The witness indicates that the hard drive examined cannot be the same drive that the MediaSentry data indicates was used to share files on the Internet. The questions that need to be addressed here are: 1) Is there any evidence that the drive is anything other than the original equipment disk drive sold with the machine? The manufacturer is actually likely to have a record of what drive (based on serial number) was sold with the machine (based on serial number). If it is the original drive, the likelihood of tampering is very low. 2) Is there any evidence that the machine was opened by anyone before the machine was impounded and the disk removed? Many manufacturers invalidate elements of warranties when a machine is opened. They are likely to be able to say. 3) Is there any evidence that the disk was reformatted and overwritten with a new copy of the operating system? Almost any disk drive recovery company could answer this question very quickly. 4) Is there any evidence that another drive (a USB or Firewire drive, for instance) was attached to the system? There should be evidence of such in the machines plug and play drivers
Slashdot Mirror
Somebody ought to label the initial "What about bans?" irrelevant and label its poster a troll. I can't do it, but this one is not only unrelated to the questions of civil liberties raised, but arguably the product of either stupidity or payola.
I have a very good friend who, despite knowing all the reasons why she shouldn't, smokes. This years "Stubbed Out" made an impression. If the effect is to get one person to stop smoking (or doing other really stupid things like igniting gasoline in florescent tubes), they server a valuable social function.
Obviously the Tesla effect is not what's interesting here. Its the direction of power to the device that is interesting. You have this big power sheet and it only sends power where there is a device that needs it.
... maybe ...
The obvious place for this power supply is your chair and bed. You can sit down with your cell phone in your pocket and have it recharge while you work at your computer.
The big deal application of this kind of technology is probably computer implants. Its hard to change the batteries of a computer that's embedded in your thigh, but now it can recharge while you sleep.
I see a big market for upholstery and bedding growing out of this device.
So I'll take my Egg McMuffin with the Florescent Green Canadian Bacon. Hmmm. I think not.
Here's an accomplishment that will create a real conundrum for the Green Party. Is the pig a symbol to be cherished, a genetic freak to be feared, or some combination.
As long as the interfaces are standard and the software can be ported from one machine to the next, there shouldn't be a problem. The latter is, of course, an argument against using Windows as the platform, but there are good Windows emulators on Linux, so all is not lost.
Not at all. There is no necessary fault in being Republican or Democrat or in the unrelated categories of being liberal or conservative. Republicans do win, and should win, elections without resorting to chicanery. Democrats do, and should do, the same.
Where, however, substantial doubt can be cast on an election outcome, due diligence should be used to resolve the issue to the extent that it is possible to do so. If there was a problem with the voting machines in this district that caused a huge number of votes to evaporate, it should be investigated. If the problem can be documented, corrective actions should be taken. That's all that this litigation is asking for.
The problem here isn't a Republican or Democratic problem. It is a problem of ensuring the election process is conducted correctly.
The problem with this judges decision is that it (in my view incorrectly) stops the due diligence process for exactly the wrong reasons. A corporations proprietary interests should NEVER trump the electoral process or throw its results into doubt.
Which is exactly why all election software should be publicly verifiable open source.
Ah, the things you don't see in preview (even with "a" and "o" so far apart on the keyboard).
It remains that making open source a requirement in voting software would be a good piece of electoral reform legislation.
Good points, but resolvable. I believe, at this point, that the only means by which we can provide full confidence in the system is for people to be able to confirm their votes after the fact. There may be no way of preventing vote buying in such a system, but I believe that the take away can be sufficiently secured as to make intimidation difficult and vote buying more readily detectable. In "There ought to be a low", below, I describe the copy as encrypted. Assuming the copy is made with a public key, the voter should be able to check their vote while still in the booth and then later (after the votes have been counted) online. They would be equally able to present the copy to a courtroom if needed, but the copy would not be human readable, immediately accessible outside of the voting booth, or accessible at any point without the keys. Such a system would not stop vote buying, which happens even without a means of verification. Most people take their votes more seriously than that, but the confirmation system would allow the identification of any machine that verifies a large number of ballots, which ought to make it easier to detect and punish vote buying. There are no technical panaceas to resolve ill behavior, but it ought to be possible for the 18,000 people who may have been disenfranchised in the election documented here to check to make sure their votes were properly counted.
It is generally much easier for us to critique new systems, like digital systems are today, than it is to critique well established systems, which often have glaring flaws of their own. We don't see the flaws in older systems because we have internalized and/or have been socialized into what is good about them and have accepted the glaring flaws of existing systems as normal, unavoidable, or offset by the positives.
There are many kinds of data insecurity. Recent demonstrations of the inherent "insecurity" of digital information focus on particular kinds of insecurity, including:
- the possibility that data can be undetectably replaced (the reason why voting machines should be secure and off network).
- the possibility that data can be undetectably miswritten (the reason why reviewing electoral source code matters).
Punch card systems (which are pseudo-digital) have been demonstrated to be open to other kinds of insecurity:Paper votes have historically been open to still other kinds of insecurity
I would argue that the mechanical machines we've been using in New York for generations resolve most of the forms of insecurity outlined above. They certainly provide a publicly verifiable ballot and a reasonably good (two level) audit trail, but they aren't perfect either, as there is no way for an individual voter to confirm that their vote was recorded correctly.
A properly done digital voting system can provide much better security than any existing system has managed. That won't happen, however, if we let the scandalously operated proprietary voting machines of the present stop us from looking for better ways to do things.
Paper ballots can dissapear into the trunk of a car and never be seen again, as reportedly happened in Jacksonville, FL in the 2000 election.
You are right, simple is better, but wouldn't it have been better if the observers for both parties could have taken a copy of the voting machine with them on a USB dongle at the end of the day? Wouldn't it be better if voters could carry a copy of their vote home with them so that it could be accounted for after the fact if necessary?
In this case, simple would be simple open source code that interpreted a simple human readable (e.g. XML) ballot specification with votes verifiable at several levels.
Pencil and paper simply provide other means of stuffing the ballot box, ways that are much harder to detect.
A good hacker who can engineer access can always get access to the software, whether open source or proprietary. Obscurity is no defense except in the courts, where hacking the code would be illegal and therefore inadmissible (and probably worse).
The big lesson of open source is that lots of eyes tend to make software more secure because concerned reviewers of the code find the flaws and, with the agreement of the large community, fix them. The security of Linux is far better than that of windows and, judging from comments about Vista, far less intrusive.
Your comment directly contradicts the article that starts the discussion.
This is a case where a Democrat is alleging voting irregularities and in which the Democratic party is allowing the Republican to take a seat in Congress despite the fact that the outcome of the election is still in the courts.
Oh, and by the way, Clinton took several shots at taking out Bin Laden (cruise missile attacks on Al Qaeda training camps in Afghanistan, as I recall). They didn't work, but they were well documented in the newspapers along with Republican assertions that Clinton was "wagging the dog".
Get your facts straight.
This is an idea that won't die, and the car companies (Ford, GM, Daimler Chrysler, Toyota, Honda, etc) have a long history of talking to the big computer companies (IBM and now Microsoft) when they explore it. If Ford went to Gates, its because he regards Gates as a premier expert. If he focused on Windows, its simply because its widely used and has a logo that will go nicely on the windows of your car. Hobbyists have been tricking out their cars with computers for years, and the DVD players that are hanging in the car of every family who takes a trip with a five year old are already turning into Video Game stations.
The good news is that, when cars start arriving in volume that have computers built in, there will quickly be:
None of this is intended to discount the very real problems associated with actually using such a computer while you are driving, but if something is going to happen eventually, better to seriously consider its implications, if only because there may be profit in doing so.
The software for electronic voting is not complicated. I've written software in the general category (survey software) myself. Good survey software should have only two parts:
The first part should be open source. There is no reason for it to be proprietary except to hide its workings or pretend that its complicated (which it isn't). It should be sufficiently stable that it can run as a ROM. If it isn't, it shouldn't be in use. If it isn't simple, moreover, than more attention has been paid to bells and whistles than to the function of the software, which should:
The result is a tally sheet that can be immediately checked at the end of voting and three levels by which the result can be double checked> The third level (a copy of the ballot cast that the voter takes with them) will allow individuals to ensure that their ballot was counted correctly.
All of which is a run up to saying that there ought to be a law that requires voting software to be open source such that every level of the process is transparent to voters.
Obviously someone who is unaware of the millions of machines that are routinely overheated by overload ... most machines running graphics intensive applications and all machines running BOINC.
Bad thinking, but wishful thinking often is.
Davis
I'm afraid it was rather long, but there are so many possibilities. Hope something in there helps.
I don't see any glaring faults in the Curriculum Vita. Doug may have some biases relative to P2P file sharing (invalid biases in my view: the first mark of a successful new medium of communication is the assertion that it is being used for pornography), but those biases hardly seem relevant to this case. His background in networking and security is at least as good as some 10,000 other people who might be called as expert witnesses, and he's willing to testify on behalf of the RIAA, which very few of those 10,000 people would be willing to do, mostly based on the well justified belief that the RIAA is using these suits to defend oligarchical control of music and video rather than defend intellectual property rights. He certainly has a background in security, networking, and computer forensics that would make him a credible witness from a technical perspective. The only fault I might point at is that it demonstrates no specific knowledge or evaluation of MediaSentry. His background and patent activity in the area of detecting and blocking P2P streams should, however, make him reasonably qualified to evaluate MediaSentry. By the way, it is probably worth asking him to evaluate MediaSentry. His evaluation may not glowing.
Dr. Jacobson's declaration is another matter entirely. It uses nine pieces of evidence, none of which are available to us here, to comment on three points in the connectivity chain that purportedly ties a computer to the Kazaa file sharing system, allegedly for the purposes of both downloading and sharing music files. Those three points are (a) the computer itself, as represented by its disk drive image, (b) the computers connection to Verizon Internet Services and, by implication, to the Internet, and (c) the detection and documentation of Internet activity associated with a specific set of IP addresses by MediaSentry. That is a major collapsing of the points of connectivity between a computer and a remote server. If I were investigating (and this is an area where I have patents) the connection points from a computer to a remote server, I'd identify quite a few more, starting with the intermediate modem, cable modem, or DSL modem and the downstream DHCP server maintained by Verizon, but this is all we have. I will return to this issue below, but there is value in addressing each of the three points in the chain that Dr. Jacobson does identify, and I'll return to this as we discuss the second issue, the computer's connection to Verizon.
First, the hard drive. The witness indicates that the hard drive examined cannot be the same drive that the MediaSentry data indicates was used to share files on the Internet. The questions that need to be addressed here are: 1) Is there any evidence that the drive is anything other than the original equipment disk drive sold with the machine? The manufacturer is actually likely to have a record of what drive (based on serial number) was sold with the machine (based on serial number). If it is the original drive, the likelihood of tampering is very low. 2) Is there any evidence that the machine was opened by anyone before the machine was impounded and the disk removed? Many manufacturers invalidate elements of warranties when a machine is opened. They are likely to be able to say. 3) Is there any evidence that the disk was reformatted and overwritten with a new copy of the operating system? Almost any disk drive recovery company could answer this question very quickly. 4) Is there any evidence that another drive (a USB or Firewire drive, for instance) was attached to the system? There should be evidence of such in the machines plug and play drivers