Why are you letting untrusted machines on your trusted networks? These mobiles should be sitting in some kind of an internal DMZ.
And if you haven't got anything between your VPN users and your internal network, you are a *moron* and you deserve to be fired on the spot for pure incompetence. Entrusting your trusted network to machines you don't administer (and which are likely home machines administered by exec./sales types) is stupid.
Your "perfectly competent" admin should be quarantining machines that are carried out of the office. Laptops like this should be sitting on a separate subnet (physical or logical), and should not be allowed on the same subnet as the desktop boxes.
There are paper systems that do not require trust of the election operator. I outlined just one such system in my post. This is the system in use for Canadian federal elections, and it works because the ballot box is observed by voters, candidates' reps and poll clerks from before the doors are opened on election day.
The box is inspected and sealed before the voting starts, the box is in plain view throughout voting, and the box is opened and counted in plain view at the polling station when voting ends.
Because the system is transparent from end-to-end, and because of the competing interests of the candidates' representatives there is almost no opportunity for ballot-stuffing.
The lists were enforced by county officials, but were produced by and had the rules for their production set by state officials. The Sec. of State set the rules for the production of the lists, then handed the lists to the county officials with instructions to scrub those names from the lists. A few counties did, actually, check the accuracy of the lists. These counties refused to use the lists.
The 90,000 people who were illegally denied registration (they were not scrubbed from the list, they were just not allowed to be added to the list) were denied registration because of actions from Bush's office, directing registration officials to deny these individuals registration - in direct contravention of a Supreme Court decision.
Which would have been fine (and was required under the law of the land) were it not for the meddling of Republican Secretary of State (And George W. Bush campaign co-chair) Harris, and Jeb Bush in setting the rules for this purge. The decision to use Soundexes, the decision to purge based on similar names, the decision to ignore middle names completely, the decision on how many points of coincidence, the decision not to include SSN's as matching criteria were required to be considered a match were decisions made by Bush and Harris.
Jeb Bush decided to _ignore_a_Supreme_Court_ruling_ and illegally deny registration to up to 90,000 individuals who had been convicted of felonies out-of-state, and had their voting rights restored in their state of origin before arriving in Florida.
So, yeah, a Democrat may have commissioned ChoicePoint to do the job, but the Republicans set the rules.
How much effort does it take to change 1000 e-ballots? How much effort does it take to add 1000 paper ballots? How likely are you to be personally apprehended if someone you thought wasn't going to vote does vote before you get there?
Go out and do your research before you post. The GNU FREE project (www.free-project.org) has been around for ages. Why don't you visit that site and see why it is that development on the project has halted in favour of education and lobbying against purely electronic voting?
Please actually go out and read something. Go to http://www.notablesoftware.com to read Dr. Mercuri's writings on the subject. Go to http://www.blackboxvoting.com to see Bev Harris' take on the very real problems with DRE/ Internet / purely electronic voting. Take a look at http://www.free-project.org to see why somebody who has actually tried to write such a system has decided to terminate the project and campaign against purely electronic voting.
The problem is nowhere near as simple as you suggest
"paper ballots are just as suspect as a computer ballot."
Please explain this completely unsupported statement.
"I suppose you could implement the concept of a vote reciept... At any point, you should be able to use that transaction number to verify your vote."
Which allows you to be sure that in a published list of votes, that your vote is marked correctly. It says nothing aboput the accuracy of the tally and tabulation. You see your vote, surrounded by a couple million other votes, which might or might not correspond to legitimate voters.
The receipt, or serial number, or comparable verification options are not only useless at preventing fraud, they damage anonymity. If a voter can choose to abandon anonymity and prove his vote, the election is further opened to vote-buying and intimidation. There is no market for votes that you can't verify were cast as sold.
"No matter how a system is designed, once the ballot becomes anonymous, you loose all tracking ability and hence leaves a large hole for hacking or rigging an election."
This statement is untrue. A paper system where:
- Ballot box is inspected before polls open by interested parties with diametrically opposed interests (namely candidates' representatives) - Ballot box is sealed in the view of candidates' reps, poll clerks, in such a way that ballots can be deposited, but not removed. - Ballot boxes kept in a location where candidates' reps and poll clerks can see them at all times during the conduct of the election. - Poll clerks receive xx number of paper ballots, individually serialled on a removable counterfoil. - Poll clerks hand out ballots sequentially and enter voters names in a record sequentially. - Poll clerks initial the counterfoil before handing out the ballot. - Voters mark the ballot, verifies that it is marked the way he wants to cast his vote. - Poll clerks confirm that the ballot came from them (their initials on the counterfoil, serial number matches one out). - Poll clerk removes counterfoil without unfolding ballot. - Voter inserts ballot into box. - At end of voting, doors are locked, ballot box is unsealed and counted by poll clerk and candidates' representatives.
Now - votes cannot be tracked to elector once in the box, and counterfoil removed. We are assured that:
1) The voter is confident that his intent is accurately indicated on the actual instrument of voting. (No MITM attack between the voter and the ballot, as is possible with DRE machines.)
2) Ballots have not been removed from the ballot box during the course of the election. (The box has been in plain view, and sealed throughout.)
3) The box has not been stuffed with multiple ballots from one source. (The box has been in plain view. The poll clerk knows how many ballots out, knows s/n's of counterfoils, will not allow a ballot he didn't hand out to go into the box)
4) The election is anonymous - we cannot verify individual voters' votes with or without their participation.
"If there's a slight cost to vote, then it will ensure that anyone who votes will have some motivation to vote and a reason for choosing one candidate over the other. That way I'd say its a good thing. "
"2) The voting machines simply record the number of votes for each candidate, and no record is created about who voted for whom"
You clearly have absolutely no understanding of electoral mechanics, or the means that can be used to trace and/or track voters. Besides which... How do you know what information the machine keeps? Have you examined its code? Its components?
"3) Election commision in India is an independent constitutional body and has been know to re-conduct the elections in voting areas with slightest hint of fraud."
All of the independance in the world isn't going to help them discover the hints of fraud in the first place. Electronic voting machines and centralized counting allow greater opportunity for fraud to go on transparently and undetectably.
"4) Each voting booth in India is allowed to have has one representative from each candidate to ensure that the other candidate does not tryto defraud the voter. This is not perfect but ensures that the fraud when it happens does not skew the result too much."
The observer role becomes utterly meaningless in a DRE election, because the real mechanics of such an election cannot be directly observed by humans. The artifact of voting cannot be observed by observers or voters, as it can in a paper election. The protection against fraud provided by observers with opposed goals in a paper election does not necessarily extend to a DRE election.
"5) The voting machines contain no OS. The code is in assembly in tamper proof chips, making it very hard to hack"
Okay, so you have no understanding of technology as well. If they are digital machines, they need an OS. They need code to interpret user input and to provide appropriate user feedback. That code may be in tamper ressistant chips, but there is no such thing as tamper proof chips. All of the tamper restistance in the world is for naught if there is a software error or back door. (Presumably you believe that assembly provides automatic bounds checking on buffers? How exactly does assembly magically protect you against bugs and back doors?)
"6) The voting machines are not linked together over a network. This implies that to tally votes the machine has to be taken to a central station where again representatives from each candidate ensure that no wholscale fraud takes place."
Centralized counting decreases reliability and increases opportunity for fraud. The chain of custody of a machine, disk, flash chip, memory cartridge, other storage device cannot be readily established, and fraud, theft or damage can occur en route between the polling station and the central counting facility. If such things occur, there is _no_ fallback error-handling in the process. There is no ready way of detecting such occurrences, much less fixing them.
"Booth capturing and rigging will now completely vanish"
Yeah. Rigging will vanish because it moves away from the thuggery and violence previously required, and toward the more sophisticated, genteel manipulation of electronics. The rigging just moves out of the public's eye.
Kids have been growing up with computers for a long time already. We have a generation through school that has been exposed to computers throughout their school life. This has encouraged some of them to think and feel confident enough to challenge the status quo.
This has encouraged the vast majority to simply use what is placed in front of them. Just as we have had a number of generations immersed in a life with cars, the vast majority of these people are not able to tinker with their cars, to modify them, or even to properly understand them.
There are a small number of people who think critically, explore and challenge. There are a vast majority who go with the flow.
Installing from source can take less time and be much easier than installing binary packages. Binary packages depend on one specific set of library versions. For many packages these dependancies are difficult to maintain, without a cascade resulting in the need to upgrade many other packages in your system. If you're building from source, though, you can build the package against whatever versions of the libraries you already have installed...
"SCO is claiming that code which they own was contributed to Linux, which means that it was not contributed by the copyright owner, which means that it was not properly put under the GPL. If you don't own code, you can't put it under the GPL."
SCO also sold a product which contained the kernel, with all of this code, licensed under the GPL. If you are the copyright holder, you can license the code under the GPL. Once you have done so, you can't back out.
So fine. Maybe they own some of the code. But they released it, distributed it and licensed it under the GPL.
"I predict that SCO will announce a Linux Licensing product. I predict that they will tell people that if they buy the license, they will be entitled to use Linux free of SCO's claims. I predict that SCO will claim that Linux users who do not purchase a license from SCO will be vulnerable to a lawsuit from SCO."
At which point, the developers who own code in the rest of the kernel will sue SCO because SCO lacks a license to modify and distribute code owned by other developers, encumbered by such restrictions.
Sec. 107. - Limitations on exclusive rights: Fair use
Notwithstanding the provisions of sections 106 and 106A, the fair use of a
copyrighted work, including such use by reproduction in copies or phonorecords or by any
other means specified by that section, for purposes such as criticism, comment, news
reporting, teaching (including multiple copies for classroom use), scholarship, or research,
is not an infringement of copyright. In determining whether the use made of a work in any
particular case is a fair use the factors to be considered shall include -
(1)
the purpose and character of the use, including whether such use is of a
commercial nature or is for nonprofit educational purposes;
(2)
the nature of the copyrighted work;
(3)
the amount and substantiality of the portion used in relation to the copyrighted
work as a whole; and
(4)
the effect of the use upon the potential market for or value of the copyrighted
work.
The fact that a work is unpublished shall not itself bar a finding of fair use if such
finding is made upon consideration of all the above factors
If you'd bothered to read the linked documents, you'd realize that this is not a patent infringement case. It is a trade secrets / copyright case. IANAL, so I don't know what level of specificity is normally in such a complaint, but they're not specific as to what code violates copyright or trade secrets. They seem to try to claim that the whole of AIX, and by consequence, the whole of IBM's UNIX knowledge is tainted by the NDA's and contractual agreements reached when IBM licensed AT&T code. This is going to be a very difficult argument for them to make at trial. Mind you, perhaps less difficult because they're opting for a jury trial - throw the dice on 12 people who can't pronounce "operating system" let alone make reasonable judgements on it.
Look. This is a proprietary algorithm which was developed by a non-cryptographer, and which hasn't been peer-reviewed. It is snake-oil until it has been exposed to the light of peer-review.
I tried to give them a few examples of companies distributing binary only drivers (NVIDIA and Rational)
Note that (at least with NVIDIA), they started distributing source for _technical_, not license issues. It was too costly for them to evaluate and ensure binary compatibility with so many versions of the kernel. So, initially, they released obfuscated source, so you could compile, but not really understand. If they were to fall under the GPL, this would not be permitted.
The company was very concerned about GPL issues and consulted a lawyer - who advised us to go for a user-space driver... Is writing a user-mode (and hence not very efficient)
From a technical perspective, if you don't *need* your driver in-kernel, don't put it there. I mean, if you're doing a storage or filesystem or ethernet driver, you usually _need_ to be in kernel. Otherwise, you usually don't... I learned this lesson by having a driver rejected for inclusion, because it didn't belong in kernel.
That said, for most of the subsystems, there are user-mode APIs that are low-level enough that you don't take a huge performance hit from going to user-mode. USB devices, for example, take a _very_ small performance hit from going user-mode.
Slashdot Mirror
RPC is used for IPC. Very stupid. :)
Why are you letting untrusted machines on your trusted networks? These mobiles should be sitting in some kind of an internal DMZ.
And if you haven't got anything between your VPN users and your internal network, you are a *moron* and you deserve to be fired on the spot for pure incompetence. Entrusting your trusted network to machines you don't administer (and which are likely home machines administered by exec./sales types) is stupid.
Your "perfectly competent" admin should be quarantining machines that are carried out of the office. Laptops like this should be sitting on a separate subnet (physical or logical), and should not be allowed on the same subnet as the desktop boxes.
There are paper systems that do not require trust of the election operator. I outlined just one such system in my post. This is the system in use for Canadian federal elections, and it works because the ballot box is observed by voters, candidates' reps and poll clerks from before the doors are opened on election day.
The box is inspected and sealed before the voting starts, the box is in plain view throughout voting, and the box is opened and counted in plain view at the polling station when voting ends.
Because the system is transparent from end-to-end, and because of the competing interests of the candidates' representatives there is almost no opportunity for ballot-stuffing.
The lists were enforced by county officials, but were produced by and had the rules for their production set by state officials. The Sec. of State set the rules for the production of the lists, then handed the lists to the county officials with instructions to scrub those names from the lists. A few counties did, actually, check the accuracy of the lists. These counties refused to use the lists.
The 90,000 people who were illegally denied registration (they were not scrubbed from the list, they were just not allowed to be added to the list) were denied registration because of actions from Bush's office, directing registration officials to deny these individuals registration - in direct contravention of a Supreme Court decision.
Which would have been fine (and was required under the law of the land) were it not for the meddling of Republican Secretary of State (And George W. Bush campaign co-chair) Harris, and Jeb Bush in setting the rules for this purge. The decision to use Soundexes, the decision to purge based on similar names, the decision to ignore middle names completely, the decision on how many points of coincidence, the decision not to include SSN's as matching criteria were required to be considered a match were decisions made by Bush and Harris.
Jeb Bush decided to _ignore_a_Supreme_Court_ruling_ and illegally deny registration to up to 90,000 individuals who had been convicted of felonies out-of-state, and had their voting rights restored in their state of origin before arriving in Florida.
So, yeah, a Democrat may have commissioned ChoicePoint to do the job, but the Republicans set the rules.
How much effort does it take to change 1000 e-ballots? How much effort does it take to add 1000 paper ballots? How likely are you to be personally apprehended if someone you thought wasn't going to vote does vote before you get there?
Go out and do your research before you post. The GNU FREE project (www.free-project.org) has been around for ages. Why don't you visit that site and see why it is that development on the project has halted in favour of education and lobbying against purely electronic voting?
Please actually go out and read something. Go to http://www.notablesoftware.com to read Dr. Mercuri's writings on the subject. Go to http://www.blackboxvoting.com to see Bev Harris' take on the very real problems with DRE/ Internet / purely electronic voting. Take a look at http://www.free-project.org to see why somebody who has actually tried to write such a system has decided to terminate the project and campaign against purely electronic voting.
The problem is nowhere near as simple as you suggest
"paper ballots are just as suspect as a computer ballot."
Please explain this completely unsupported statement.
"I suppose you could implement the concept of a vote reciept... At any point, you should be able to use that transaction number to verify your vote."
Which allows you to be sure that in a published list of votes, that your vote is marked correctly. It says nothing aboput the accuracy of the tally and tabulation. You see your vote, surrounded by a couple million other votes, which might or might not correspond to legitimate voters.
The receipt, or serial number, or comparable verification options are not only useless at preventing fraud, they damage anonymity. If a voter can choose to abandon anonymity and prove his vote, the election is further opened to vote-buying and intimidation. There is no market for votes that you can't verify were cast as sold.
"No matter how a system is designed, once the ballot becomes anonymous, you loose all tracking ability and hence leaves a large hole for hacking or rigging an election."
This statement is untrue. A paper system where:
- Ballot box is inspected before polls open by interested parties with diametrically opposed interests (namely candidates' representatives)
- Ballot box is sealed in the view of candidates' reps, poll clerks, in such a way that ballots can be deposited, but not removed.
- Ballot boxes kept in a location where candidates' reps and poll clerks can see them at all times during the conduct of the election.
- Poll clerks receive xx number of paper ballots, individually serialled on a removable counterfoil.
- Poll clerks hand out ballots sequentially and enter voters names in a record sequentially.
- Poll clerks initial the counterfoil before handing out the ballot.
- Voters mark the ballot, verifies that it is marked the way he wants to cast his vote.
- Poll clerks confirm that the ballot came from them (their initials on the counterfoil, serial number matches one out).
- Poll clerk removes counterfoil without unfolding ballot.
- Voter inserts ballot into box.
- At end of voting, doors are locked, ballot box is unsealed and counted by poll clerk and candidates' representatives.
Now - votes cannot be tracked to elector once in the box, and counterfoil removed. We are assured that:
1) The voter is confident that his intent is accurately indicated on the actual instrument of voting. (No MITM attack between the voter and the ballot, as is possible with DRE machines.)
2) Ballots have not been removed from the ballot box during the course of the election. (The box has been in plain view, and sealed throughout.)
3) The box has not been stuffed with multiple ballots from one source. (The box has been in plain view. The poll clerk knows how many ballots out, knows s/n's of counterfoils, will not allow a ballot he didn't hand out to go into the box)
4) The election is anonymous - we cannot verify individual voters' votes with or without their participation.
Would you call the balloting system(s) in the United States conventional? Is a Supreme Court split on party lines the proper electorate?
"If there's a slight cost to vote, then it will ensure that anyone who votes will have some motivation to vote and a reason for choosing one candidate over the other. That way I'd say its a good thing. "
Google for "Jim Crow" "poll tax"
"2) The voting machines simply record the number of votes for each candidate, and no record is created about who voted for whom"
You clearly have absolutely no understanding of electoral mechanics, or the means that can be used to trace and/or track voters. Besides which... How do you know what information the machine keeps? Have you examined its code? Its components?
"3) Election commision in India is an independent constitutional body and has been know to re-conduct the elections in voting areas with slightest hint of fraud."
All of the independance in the world isn't going to help them discover the hints of fraud in the first place. Electronic voting machines and centralized counting allow greater opportunity for fraud to go on transparently and undetectably.
"4) Each voting booth in India is allowed to have has one representative from each candidate to ensure that the other candidate does not tryto defraud the voter. This is not perfect but ensures that the fraud when it happens does not skew the result too much."
The observer role becomes utterly meaningless in a DRE election, because the real mechanics of such an election cannot be directly observed by humans. The artifact of voting cannot be observed by observers or voters, as it can in a paper election. The protection against fraud provided by observers with opposed goals in a paper election does not necessarily extend to a DRE election.
"5) The voting machines contain no OS. The code is in assembly in tamper proof chips, making it very hard to hack"
Okay, so you have no understanding of technology as well. If they are digital machines, they need an OS. They need code to interpret user input and to provide appropriate user feedback. That code may be in tamper ressistant chips, but there is no such thing as tamper proof chips. All of the tamper restistance in the world is for naught if there is a software error or back door. (Presumably you believe that assembly provides automatic bounds checking on buffers? How exactly does assembly magically protect you against bugs and back doors?)
"6) The voting machines are not linked together over a network. This implies that to tally votes the machine has to be taken to a central station where again representatives from each candidate ensure that no wholscale fraud takes place."
Centralized counting decreases reliability and increases opportunity for fraud. The chain of custody of a machine, disk, flash chip, memory cartridge, other storage device cannot be readily established, and fraud, theft or damage can occur en route between the polling station and the central counting facility. If such things occur, there is _no_ fallback error-handling in the process. There is no ready way of detecting such occurrences, much less fixing them.
Yeah. Rigging will vanish because it moves away from the thuggery and violence previously required, and toward the more sophisticated, genteel manipulation of electronics. The rigging just moves out of the public's eye.
Kids have been growing up with computers for a long time already. We have a generation through school that has been exposed to computers throughout their school life. This has encouraged some of them to think and feel confident enough to challenge the status quo.
This has encouraged the vast majority to simply use what is placed in front of them. Just as we have had a number of generations immersed in a life with cars, the vast majority of these people are not able to tinker with their cars, to modify them, or even to properly understand them.
There are a small number of people who think critically, explore and challenge. There are a vast majority who go with the flow.
Installing from source can take less time and be much easier than installing binary packages. Binary packages depend on one specific set of library versions. For many packages these dependancies are difficult to maintain, without a cascade resulting in the need to upgrade many other packages in your system. If you're building from source, though, you can build the package against whatever versions of the libraries you already have installed...
You apparently have _no_ idea about what the job market is for PhD's who wish to continue and be hired in academe in order to continue research.
"SCO is claiming that code which they own was contributed to Linux, which means that it was not contributed by the copyright owner, which means that it was not properly put under the GPL. If you don't own code, you can't put it under the GPL."
SCO also sold a product which contained the kernel, with all of this code, licensed under the GPL. If you are the copyright holder, you can license the code under the GPL. Once you have done so, you can't back out.
So fine. Maybe they own some of the code. But they released it, distributed it and licensed it under the GPL.
"I predict that SCO will announce a Linux Licensing product. I predict that they will tell people that if they buy the license, they will be entitled to use Linux free of SCO's claims. I predict that SCO will claim that Linux users who do not purchase a license from SCO will be vulnerable to a lawsuit from SCO."
At which point, the developers who own code in the rest of the kernel will sue SCO because SCO lacks a license to modify and distribute code owned by other developers, encumbered by such restrictions.
Sec. 107. - Limitations on exclusive rights: Fair use
Notwithstanding the provisions of sections 106 and 106A, the fair use of a
copyrighted work, including such use by reproduction in copies or phonorecords or by any
other means specified by that section, for purposes such as criticism, comment, news
reporting, teaching (including multiple copies for classroom use), scholarship, or research,
is not an infringement of copyright. In determining whether the use made of a work in any
particular case is a fair use the factors to be considered shall include -
(1)
the purpose and character of the use, including whether such use is of a
commercial nature or is for nonprofit educational purposes;
(2)
the nature of the copyrighted work;
(3)
the amount and substantiality of the portion used in relation to the copyrighted
work as a whole; and
(4)
the effect of the use upon the potential market for or value of the copyrighted
work.
The fact that a work is unpublished shall not itself bar a finding of fair use if such
finding is made upon consideration of all the above factors
If you'd bothered to read the linked documents, you'd realize that this is not a patent infringement case. It is a trade secrets / copyright case. IANAL, so I don't know what level of specificity is normally in such a complaint, but they're not specific as to what code violates copyright or trade secrets. They seem to try to claim that the whole of AIX, and by consequence, the whole of IBM's UNIX knowledge is tainted by the NDA's and contractual agreements reached when IBM licensed AT&T code. This is going to be a very difficult argument for them to make at trial. Mind you, perhaps less difficult because they're opting for a jury trial - throw the dice on 12 people who can't pronounce "operating system" let alone make reasonable judgements on it.
Look. This is a proprietary algorithm which was developed by a non-cryptographer, and which hasn't been peer-reviewed. It is snake-oil until it has been exposed to the light of peer-review.
And if his email is more likely to work under one operating system, then he does care which OS he uses.
Windows crashing is almost always due to bad hardware.
Cop-out. I have worked with a number of machines which behaved very poorly under Windows, but very well under Linux.
By the way a lawyer spending $2k isn't that much. They probably make it back on their first case.
And if they don't have to spend that $2k on software, they now have $2k more that they can spend on:
Manufacturing Consent: The Political Economy of the Mass Media - Edward S. Herman, Noam Chomsky - ISBN 0375714499.
This is the 2002 edition - the book was initially published in 1988, and its model seems to have held up incredibly well.
I tried to give them a few examples of companies distributing binary only drivers (NVIDIA and Rational)
Note that (at least with NVIDIA), they started distributing source for _technical_, not license issues. It was too costly for them to evaluate and ensure binary compatibility with so many versions of the kernel. So, initially, they released obfuscated source, so you could compile, but not really understand. If they were to fall under the GPL, this would not be permitted.
The company was very concerned about GPL issues and consulted a lawyer - who advised us to go for a user-space driver... Is writing a user-mode (and hence not very efficient)
From a technical perspective, if you don't *need* your driver in-kernel, don't put it there. I mean, if you're doing a storage or filesystem or ethernet driver, you usually _need_ to be in kernel. Otherwise, you usually don't... I learned this lesson by having a driver rejected for inclusion, because it didn't belong in kernel.
That said, for most of the subsystems, there are user-mode APIs that are low-level enough that you don't take a huge performance hit from going to user-mode. USB devices, for example, take a _very_ small performance hit from going user-mode.
Linus is not the freakin' copyright holder. He is the copyright holder of a significant portion of the code, but not all of it.