|
|
 Up |
|
|
  |
Author: presidentbyamendmentpresidentbyamendment Date: Dec 9, 2007 11:09
Hi Forth-ward looking folk!
Every several years I should mention that the 3-stack machine I stuck
into the Linux kernel as an interpreter into kernelspace is in cLIeNUX/
interim, H3rL, Hohensee's 3-ring Linux. That's on linux01.gwdg.de/
~rhohen .
The reason I dropped by is to mention that I've written a 1000-line
Bash script that walks John Q. Public through the process of editing
thier paper ballot for an election. It prints a receipt with a UID
generated by VUID=$RANDOM$RANDOM$RANDOM$RANDOM, and that receipt is
tied to the ballot and a tally file of tab-delimited records that can
be cat'ed into a national tally.
validvote runs on Linux. I've carved Fedora down to a lean mean 1.9
gig. validvote could just as easily be a bootable Forth, which would
perhaps make it a bit less scary from a security point of view. I
believe elections are open source's killer app. I'm not so sure Bash
is the killer language.
The average Enth user could do a Forth validvote in a couple hours.
validvote has been very well received by those who have seen it. Very
well. There is a need for such a thing, to say the least.
|
| Show full article (1.98Kb) |
|
| | 38 Comments |
|
|
Author: Bernd PaysanBernd Paysan Date: Dec 9, 2007 12:24
presidentbyamendment wrote:
> The reason I dropped by is to mention that I've written a 1000-line
> Bash script that walks John Q. Public through the process of editing
> thier paper ballot for an election. It prints a receipt with a UID
> generated by VUID=$RANDOM$RANDOM$RANDOM$RANDOM, and that receipt is
> tied to the ballot and a tally file of tab-delimited records that can
> be cat'ed into a national tally.
Elections must be anonymous and done in secret. If you allow people to take
out a receipt which can be verified by everyone, you open up the can of
worms of vote trading. I give some poor guys ten dollars each to vote for
Gush, and perhaps you give some other guys ten dollars each to vote for
Bore. Or whatever. Payment depends on verified receipt. Votes have to be
shuffled around, since if you are able to order them, you can secretly
count who's going into the voting cabine, and then verify if the "right"
vote was achieved. If you don't like paying, violence has been found
useful, as well.
|
| Show full article (1.56Kb) |
|
| | no comments |
|
|
Author: Jerry AvinsJerry Avins Date: Dec 9, 2007 13:14
Bernd Paysan wrote:
...
> I believe paper and pencil are the right tool for elections. It must be
> possible for virtually everyone to understand the voting process, and to
> verify that everything has been done correctly.
Pencil, no, Something ineradicable.
Jerry
--
Engineering is the art of making what you want from things you can get.
¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯¯
|
| |
| no comments |
|
|
Author: Bernd PaysanBernd Paysan Date: Dec 10, 2007 01:07
John Passaniti wrote:
> Jerry Avins wrote:
>> Pencil, no, Something ineradicable.
Yes, normally we use wax crayons which can't be erased.
> And ordinary paper-- no. You need paper that can clearly show if any
> attempt has been made to edit the vote after it was cast.
Well, wax sticks quite good to ordinary recycled paper ;-).
Another important point is that you can verify that the paper ballot is
genuine, so that nobody can create his own fake ballot papers, and secretly
exchange some hundred self-fabricated votes with the real ones. Either that
or you monitor the counting process well enough to reduce this likelyhood.
|
| Show full article (3.11Kb) |
| no comments |
|
|
Author: Anton ErtlAnton Ertl Date: Dec 12, 2007 02:28
John Passaniti JapanIsShinto.com> writes:>Bernd Paysan wrote:
>> All this happens in public and can be inspected by anyone who cares. That's
>> the whole point behind it - no secrets, no technology that is only
>> understood by less than 1%% of the population. For sure, it depends on
>> people who care about the voting process, but that's a fundamental point in
>> democracy, anyway: If nobody cares, it won't work.
>
>I don't think it's a matter of caring. You're ignoring some very big
>points by saying "it can be inspected by anyone who cares." Our last
>couple big elections in the United States had people who cared very much
>and were passionate about inspecting the process. But they were stopped
>in all sorts of ways-- procedurally, legally, physically and by in some
>cases...
|
| Show full article (2.68Kb) |
| no comments |
|
|
Author: hohenseehohensee Date: Jan 22, 2008 10:11
On Dec 9 2007, 3:24 pm, Bernd Paysan wrote:
> presidentbyamendment wrote:
>> The reason I dropped by is to mention that I've written a 1000-line
>> Bash script that walks John Q. Public through the process of editing
>> thier paper ballot for an election. It prints a receipt with a UID
>> generated by VUID=$RANDOM$RANDOM$RANDOM$RANDOM, and that receipt is
>> tied to the ballot and a tally file of tab-delimited records that can
>> be cat'ed into a national tally.
>
> Elections must be anonymous and done in secret. If you allow people to take
> out a receipt which can be verified by everyone, you open up the can of
> worms of vote trading. I give some poor guys ten dollars each to vote for
> Gush, and perhaps you give some other guys ten dollars each to vote for
> Bore. Or whatever. Payment depends on verified receipt. Votes have to be
> shuffled around, since if you are able to order them, you can secretly
> count who's going into the voting cabine, and then verify if the "right"
> vote was achieved. If you don't like paying, violence has been found
> useful, as well.
>
>>validvoteruns on Linux. I've carved Fedora down to a lean mean 1.9 ...
|
| Show full article (1.86Kb) |
| no comments |
|
|
Author: pml540114pml540114 Date: Jan 23, 2008 05:22
hohensee wrote:
> On Dec 9 2007, 3:24�pm, Bernd Paysan wrote:
>> presidentbyamendment wrote:
>>> The reason I dropped by is to mention that I've written a 1000-line
>>> Bash script that walks John Q. Public through the process of editing
>>> thier paper ballot for an election. It prints a receipt with a UID
>>> generated by VUID=$RANDOM$RANDOM$RANDOM$RANDOM, and that receipt is
>>> tied to the ballot and a tally file of tab-delimited records that can
>>> be cat'ed into a national tally.
>>
>> Elections must be anonymous and done in secret. If you allow people to take
>> out a receipt which can be verified by everyone, you open up the can of
>> worms of vote trading. I give some poor guys ten dollars each to vote for
>> Gush, and perhaps you give some other guys ten dollars each to vote for
>> Bore. Or whatever. Payment depends on verified receipt. Votes have to be
>> shuffled around, since if you are able to order them, you can secretly
>> count who's going into the voting cabine, and then verify if the "right"
>> vote was achieved. If you don't like paying, violence has been found
>> useful, as well.
>> ...
|
| Show full article (2.85Kb) |
| no comments |
|
|
Author: hohenseehohensee Date: Jan 24, 2008 09:03
> Jerry Avins wrote:
>> Pencil, no, Something ineradicable.
>
> And ordinary paper-- no. You need paper that can clearly show if any
> attempt has been made to edit the vote after it was cast. This includes
> the case of people not choosing to vote in certain races. If I choose
> not to vote in some races on the ballot, I don't want others helpfully
> filling in the blanks.
>
> The appeal of paper and pencil is that voters understand it. But
> understanding doesn't mean they trust it. If I'm using such a system,
> how do I know that each piece of paper was counted? How do I know it
> was counted correctly? How do I know that the paper will be stored
> properly so that challenges to the count can be performed later?
>
> Hohensee's vote counting system is as trivial, insecure, and painfully
> naive as it was the last time he brought this up in comp.lang.forth.
> But, I can imagine features of such a system that could not only get ...
|
| Show full article (1.23Kb) |
| no comments |
|
|
Author: John PassanitiJohn Passaniti Date: Jan 27, 2008 06:18
hohensee wrote:
> And I invite the John Passanitis of the world to try to defeat my
> system.
I've done this in the past. I showed that the various dependencies you
had (bash, Linux, etc.) could themselves be compromised. Then, after
reviewing the code, I pointed out several deep flaws in the design that
could *easily* lead to vote tampering. I'm too lazy to look up my past
messages on the subject, but you're free to do so.
If you've made changes since the last time, I would be happy to review
your software and provide to you what I fearlessly predict based on the
quality of what I saw previously: A slew of both real and theoretical
ways the system can be attacked.
So, if that's what you want, please point me to the current source code
and I will be happy to rip through it and illustrate the problems it
has. Simply point me to where I can download the code. Or if it's no
longer distributed online, you can email me the code.
|
| |
| no comments |
|
|
|
|
|
Author: Albert van der HorstAlbert van der Horst Date: Jan 27, 2008 14:49
In article <479c92d8$0$1103$4c368faf@ roadrunner.com>,
John Passaniti JapanIsShinto.com> wrote:>hohensee wrote:
>> And I invite the John Passanitis of the world to try to defeat my
>> system.
>
>I've done this in the past. I showed that the various dependencies you
>had (bash, Linux, etc.) could themselves be compromised. Then, after
>reviewing the code, I pointed out several deep flaws in the design that
>could *easily* lead to vote tampering. I'm too lazy to look up my past
>messages on the subject, but you're free to do so.
>
>If you've made changes since the last time, I would be happy to review
>your software and provide to you what I fearlessly predict based on the
>quality of what I saw previously: A slew of both real and theoretical
>ways the system can be attacked.
>
>So, if that's what you want, please point me to the current source code
>and I will be happy to rip through it and illustrate the problems it
>has. Simply point me to where I can download the code. Or if it's no ...
|
| Show full article (3.94Kb) |
| no comments |
|
|
|
|
