Slashdot Mirror
E-Voting Glitch: 19,000 Voters, 144,000 Votes
nick_davison writes "The Indianapolis Star is reporting the latest case of 'interesting' E-voting results. Tuesday's Boone County election, using MicroVote software returned 144,000 votes from 19,000 registered voters. After much panicking and tracking down the bug, the actual number of votes turned out as 5,352. With yet another mistake, does anyone still trust closed-source electronic voting?"
I remember at our last national election, the voting was simple - make an X on a ballot and put it in the voting box.
I have to wonder, with all these punch cards, evote, and other problems - why don't they just stick to plain old pen & paper ballots? I mean if you can't figure those out, chances are you'll end up just stuffing your ballot into the funny "circular" ballot box anyways!
A lengthy collaboration between the county's information technology director and advisers from the MicroVote software producer ... showed just 5,352 ballots
So an IT director and a number of flunkies have rewritten the results of an election.
How do the good people of Boone County know that the new answer is correct? Because it's less than the number of actual voters? How can they trust the result of that election at all? And why should those too young to vote until next time bother to vote when next time comes around?
What matters is an accurate count. Why oh why is this so difficult? Press a button, tally a vote. Next voter please. Why is this even still being discussed??? Maybe I'm dense, but I just don't get it.
When software is used to impliment a matter of law, the public must have an absolute right and need to review such software, even before one speaks of issues of software freedom. We don't make closed source or "secret" laws in this country, ie, laws that effect the public in general, and that the public is not permitted to know or examine, but yet will be held accountable to. We don't have anonymous or secret agencies enforcing laws and arresting people, ie, a secret police force. Yet, for reasons I cannot fathom, we now permit machinary with no public means of review to impliment laws, such as voting. No democracy can exist where voting is a secret or unaccountable process to the public that participates in it.
Sod that.
With yet another mistake, does anyone trust electronic voting full stop?
(I think that Open Source might be better, but to the majority of voters, electronic voting is the same thing irrespective of how visible the code is - and quite frankly, even with peer review on open coude this sort of bug might still happen)
Avantslash - View Slashdot cleanly on your mobile phone.
So 19,000 voters produced 144,000 votes. That's obviously an error, and was caught and corrected. What you really need to worry about are the little errors; if the votes are off by 500 or 1000 how are you going to know?
...
I mean really, how difficult can this be. Lots of people vote, you add up the totals, we're not talking rocket science here. When was the last time your local ATM machine gave you $1500 instead of the $50 that you asked for. Doesn't happen too often right? Maybe it's because the banks are damned sure they're not going to give their money away. It's a pity the people in charge don't take democracy that seriously.
Skiing? Check out The Independant Skiers Portal
Yeah, the IT director and the software provider "fixing" the problem is a little bit disconcerting.
LordBodak's journal.
With our new patent-pending macro-vote system, you too can auto-vote most of your constituents in a single mouse click. And you can do it as many times as you want!
Macro-vote, for a macro generation!
Simon.
Physicists get Hadrons!
"With yet another mistake, does anyone still trust closed-source electronic voting?"
This infers that open source == no mistakes. That's simply not true. It just means that there *may* be less mistakes as theoretically more people look at it. Think SendMail... that's open source, widely used, but that sure has had plenty of "mistakes".
Read reviews of shopping cart software
even with peer review on open coude this sort of bug might still happen
But in that case we at least get to see the bug *and* the fix. Now someone has 'fixed' the count and but he could just as well have done that by inserting some hardcoded reasonable looking numbers.
Comment removed based on user account deletion
Ok, not really.
With yet another mistake, does anyone still trust closed-source electronic voting?
Open source, closed source, it does not matter. Open source is not a cure for solid software development practices, and open source is not a synonym for solid software development practices. Likewise "closed-source" does not equate to poor practices.
One of the strengths of open source is the price. Free software probably means more people are using it than would otherwise, so the software is being tested more, and the pool of people available to fix bugs is also larger. This works for software that is generally useful, but consider voting software. Who is going to install the full voting suite (voting software is much more than a voting terminal) and then hold mock elections in their home? Granted, the importance of such software may bring out more people willing to try the software but you are still relying on people to do this in their leisure time.
The "many eyes" argument is merely a shotgun approach to quality control. What is needed is strong leadership implemeting a plan which includes rigorous and ongoing testing. Open source does not guarantee this any more than closed source guarantees its absence.
The software was released before it was ready. That's obvious. It seems to me that a closed source shop would be theoretically better positioned to meet an immutable deadline (such as an election date). At least when you own your employees you can mandate overtime and crack the whip harder. When the software is open source you cannot enter "crunch mode" and make the scattered developers put in long hours.
The fault was not in the development model but in the failure of the project leadership.
E-Voting Glitch: 19,000 Voters, 144,000 Votes
I hate the word "glitch", I really do.
It's an evasion, a pathetic euphemism.
What it really means is "bad programming", "fucked up", "profoundly fucked up", etc.
-kgj
-kgj
What is the ridiculuous complexity making these things so easy to fcuk up?
Combine it perhaps with a bar code scanner so that every individual can have a street bar code. Add a few simple checks like no more bar codes are counted for a paricular street than were issued.
I still don't see where this becomes a complex task compared to existing systems. Most of the components needed to build a system already existing.
Some one please tell me what I am missing.
As for the open source/free software issue. Perhaps the solution is that the requirements for the system should be published so that anyone can right something to conform. (Oh that's like having open standards).
Quality of code and open vs closed source have nothing to do with each other. There is plenty of crap open source code out there.
We don't want to have to pay someone to tally all the votes.
If your vote is so important to why don't volunteers count the votes? Several states, Texas example, require a human readable ballot. Smaller cities may use hand counts. Most large cities use a machine/human readable "scantron" type ballots. They mark the ballots with a permanent ink marker. Marking more than one selection for the same race invalidates only the section of the ballot for that race. IF you notice you made a mistake you can get a fresh ballot. The spoiled ballot is destroyed while you watch. Observers from all parties can watch the election judges (the people that issue the ballots, destroy the miss-marked ballots and watch you put your ballot in the box).
In Europe and Canada most countries require a paper ballot. They limit the number of voters assigned to each polling place so the votes can be counted and certified within a few hours of the close of the polls. Usually they have next day official results. It does require lots of people to complete the process but most are volunteers.
Get off the open/closed source debate already. If you use electronic voting, you open the door to electronic voting fraud. Open source is helpful in this regard, but not as effective as keeping to paper voting. Think about it. You can pay people to commit fraud anyway, but the cost goes up with number of votes altered/subtituted/whatever. With electronic voting, one guy can automate the fraud process with much greater effect. You raise the efficiency of the fraud as well as the voting.
People will argue the supposed cost and efficiency advantages of e-voting. Think about the cost of counting YOUR ONE VOTE and compare that to what YOU PAY IN TAXES each year - then tell me it's expensive. It's been working fine for over 200 years, there is little to gain from changing and everything to lose.
There's a point to be made here.
The fact of the matter is that open source software will do very little to help the issue of the untrustworthiness of electronic voting.
Simply put, being able to read the source code does you no good if you can't be sure that the binary that the voting machine is running was compiled from that code.
With a Linux distro, if I for some reason suspect Red Hat may be compiling back doors into xdm or login, I can go somewhere else. If I don't trust anybody, I can compile the damn thing myself and put it on my computer.
These machines, open source or not, are going to be provided by a company like Diebold. Do you trust them, even if they have to give you a copy of some source code which may or may not be the source code that they used in their voting machines? Are you going to be able to browse the source code on the very voting machine you're using? Are you going to be given the compiler flags used to create the binary so you can re-create it yourself, and access to the voting machine's disk so you can compare them?
It is necessary that any electronic voting system be open source, as a matter of duty to the public. It is not, however, sufficient.
The enemies of Democracy are
True. However, the idea is to avoid that sort of thing unless it is truly necessary, since even though there are good reasons to keep the details of military and espionage spending secret, the secrecy can be abused and used to hide unethical and even illegal actions. It's best to keep government activity public by default and only maintain secrecy if there is a compelling reason to do so.
There was a minor fault in the vote rigging module. It's since been corrected. Move along, nothing to see here.
I'm trying to teach myself to set people on fire with my mind... Is it hot in here?
I would hope that any company that supplied software for something like counting votes would have to provide evidence of a complete testing procedure that would catch problems like this.
I mean, automated testing of a voting system can't be hard. Build yourself a little network of voting machines in the office, write a bunch of scripts that enter a certain pattern of votes and ensure the correct results come out the other end. Make sure your scripts perform a wide range of possible voting patterns, and do all the 'odd' things your users might do (try to vote twice, mash the keypad with their palm etc).
Or am I being terribly naive about the way the software industry does things?
It's not just that 19000 voters produced 144000 votes; it's that 19000 voters produced 5352 BALLOTS that produced 144000 votes.
Obviously, this was intended as the Chicago or Baldwin release of the software.
//Information does not want to be free; it wants to breed.
>We don't make closed source or "secret" laws in
>this country, ie, laws that effect the public
>in general, and that the public is not
>permitted to know or examine, but yet will be
>held accountable to. We don't have anonymous or
>secret agencies enforcing laws and arresting
>people, ie, a secret police force.
are you serious? two words my friend: Patriot. Act.
get your memory checked now: http://www.thememoryhole.org
> We don't have anonymous or secret agencies enforcing laws and arresting people, ie, a secret police force.
Yes, as a matter of fact, in the U.S.A., we do have this.
And it's so much easier when you can rig an election.
Why does it matter what the source says - the ballots should be printed out on standard form from the e-voting machine. Vote, print the ballots out, done.
If the electronic number doesn't verify, scan or count the paper ballots by hand.
Back in the days when votes were counted by hand {or today, in countries where they stil are} the whole process was transparent.
If your country uses electronic voting, you should write to your representative and point out the necessity of opening up the process. Specifically, the need for the public to be able to examine mechanical drawings and software source code. Public scrutiny over the democratic process is more important than any corporate secret.
Je fume. Tu fumes. Nous fûmes!
Fine. I will take this troll.
I don't think the author or anyone else is saying "choose OSS because it's bug free". I believe the point to be made is that Open-source can be inspected by anyone since it's available to the public.
I for one do not trust proprietary software for voting. Government should be transparent and so should the software used to elect those buffoons into office.
And I will go so far to say that not only should the software be OSS. I should be able to download the voting data and run my own analysis of the past election.
Whoa, whoa, whoa, whoa, whoa, whoa, whoa, whoa, whoa, whoa, whoa, whoa. Lois, this isn't my Batman glass. - Peter
When you cheat an ATM machine, the ATM machine owner loses money. When you cheat an electronic voting machine, the machine owner may have no stake in the results or may even be benefitted by your action.
When an ATM machine cheats you, you know it, often immediately. When a voting machine cheats you, in a secret ballot system with the simplistic unauditable voting machines we use now, you never find out.
You got it wrong. Open-source doesn't mean Linux-style development. You can have a company-regulated project, but still open to the citizens wishing to peek at the code.
Gotta love them. Good thing they do not have an ulterior motive.
http://saveie6.com/