Slashdot Mirror

Yep, it's been used over here, and runs on Linux live CDs. http://www.softimp.com.au/evacs/index.html

There's a Wired article here: http://www.wired.com/techbiz/media/news/2003/11/61 045

Here in Australia we have a system that works, and has been used already.

http://www.softimp.com.au/index.php?id=evoting

This open-source system was developed by a number of well known names in the open source community - including - Andrew Tridgell (Samba), Martin Pool (Apache), and Rusty Russell (ip-tables / netfilter).

All elections for the ACT government in Australia are now run using this system. Votes are lodged either at an eVACS terminal or - if lodged on paper ballot sheets - are manually entered into the electronic system for counting. That is - all votes end up in electronic form before counting / preference distribution is done automatically by computer.

more info and source code from the electoral office and the government recommends continued use following a full review after the last election.

There are a couple of factors that meant electronic counting / voting were going to come sooner rather than later in the ACT: the useful base of some well regarded open source leaders + the ACTs difficult Hare-Clark preference distribution scheme (allowing the part of your vote unnecessary to elect your prefered candidate to go on and help elect your next prefered candidate).

+laughing at US politics paragraph+ Obviously the $200,000 cost of development of such an open, accurate, and secure system is clearly not high enough to give US governments' bank rollers the belief they are getting value for money from their political donations! Maybe Halliburton can develop such a system for use in the US for a billion or so?

This open-source system was developed by a number of well known names in the open source community - including - Andrew Tridgell (Samba), Martin Pool (Apache), and Rusty Russell (ip-tables / netfilter).

All elections for the ACT government in Australia are now run using this system. Votes are lodged either at an eVACS terminal or - if lodged on paper ballot sheets - are manually entered into the electronic system for counting. That is - all votes end up in electronic form before counting / preference distribution is done automatically by computer.

more info and source code from the electoral office and the government recommends continued use following a full review after the last election.

There are a couple of factors that meant electronic counting / voting were going to come sooner rather than later in the ACT: the useful base of some well regarded open source leaders + the ACTs difficult Hare-Clark preference distribution scheme (allowing the part of your vote unnecessary to elect your prefered candidate to go on and help elect your next prefered candidate).

+laughing at US politics paragraph+ Obviously the $200,000 cost of development of such an open, accurate, and secure system is clearly not high enough to give US governments' bank rollers the belief they are getting value for money from their political donations! Maybe Halliburton can develop such a system for use in the US for a billion or so?

We geeks need to contribute to the open source voting software efforts!!

There are only two very early stage projects for the US market:

http://www.openvotingconsortium.org/

http://www.softimp.com.au/index.php?id=evacs

I'm trying to help out openvotingconsortium.org and am reading up on the other one which I just found out about.

What are you doing??

Why doesn't some company start up and open e-voting machine business?

This open-source system was developed by a number of well known names in the open source community - eg - Andrew Tridgell (Samba), Martin Pool (Apache), and Rusty Russell (ip-tables / netfilter).

This system is in active use. All elections for the ACT government in Australia are now run using this system. Votes are lodged either at an eVACS terminal or - if lodged on paper ballot sheets - are manually entered into the electronic system for counting. That is - all votes end up in electronic form before counting / preference distribution is done automatically by computer.

Obviously the $200,000 cost of development of such an open, accurate, and secure system is clearly not high enough to give US governments' bank rollers the belief they are getting value for money from their political donations! Maybe Halliburton can develop such a system for use in the US for a billion or so?

more info and source code from the electoral office and recommendation for continued use.

Why doesn't some company start up and open e-voting machine business?

This open-source system was developed by a number of well known names in the open source community - eg - Andrew Tridgell (Samba), Martin Pool (Apache), and Rusty Russell (ip-tables / netfilter).

This system is in active use. All elections for the ACT government in Australia are now run using this system. Votes are lodged either at an eVACS terminal or - if lodged on paper ballot sheets - are manually entered into the electronic system for counting. That is - all votes end up in electronic form before counting / preference distribution is done automatically by computer.

Obviously the $200,000 cost of development of such an open, accurate, and secure system is clearly not high enough to give US governments' bank rollers the belief they are getting value for money from their political donations! Maybe Halliburton can develop such a system for use in the US for a billion or so?

more info and source code from the electoral office and recommendation for continued use.

Replying to myself, here are some links relating to open source voting technology:

Project to provide open source voting to California

Australian elections done right with open source voting software

1. Privacy of voter hmm. not a software problem. plywood works fine. cardboard too, in a pinch
2. Authenticity of voter i made a suggestion
3. Avoidance of coercion not a problem for software. possible solutions? plywood, again, works. people not letting themselves being coerced. neutral parties supervising (castro suggested cuban supervisors after the last presidential election fiasco. i actually agree with him, maybe just not all one nationality)
4. Empty ballot box at start of polling this is a software problem? i would hope that a ballot box would be completely emptied at the end of the previous election. voters have a pesky preference of having their votes count(ed). or maybe that's just me
5. Security of ballot papers this company sells evoting machines. i would think the point of that is to get rid of paper. why reintroduce paper into the mix? rather, why try to get rid of paper in the first place, only to reintroduce it? oh ... get rid of *most* of the paper. ok, that's understandable
6. One vote per person ok. the machine could punch a hole in some card mailed to registered voter. sort of like the read-only tab on floppies

Not As Such.

The data about the ACT (Australian Capital Territory) election, the open-source software, the open-source operating system it ran on, and the open-source compiler that compiled it all is available via the makers, Software Improvements Pty Ltd.

The independant review, lab results of testing, and report on how it all worked in practice are available through the ACT Electoral Commission.

Not so much a well-oiled Chevy as a Holden Monaro which in the US is called a Pontiac GTO. As the GM site says, What Makes an American Legend? But it doesn't say the answer - good Aussie Engineering. Oi! Oi! Oi!

here

eVACS is open source and has already been used in some Australian elections. I did a research project on electronic voting systems last semester, and eVACS seemed to be the best current system. IMO, it would be much better with a voter-verified paper trail, though.

VoteHere a company that makes software to implement a particular voting crytographic scheme is the second outfit to release their source (the first was OVC).

The Third, I think.
Software Improvements did so in 2001. Source Code available through the ACT Electoral Commission. Open-source OS, Open-Source compiler, and runs on non-proprietary hardware too.

Our system printed receipts for votes, had internationalization, allowed for various layouts of the ballot on screen, and made no assumptions as to how many candidates and parties there were. The ballot was configurable from a text file, and the computer could be switched off at any point during the voting process, and you could tell if the vote was counted or not... well there was an infinitesimally small chance that the power could go at just the right time... and the vote was counted before it was logged on the local machine. You'd probably have about a 1ms window to hit the power if you were trying to sabotage the system.

Not good enough. That 1ms window means the system is unusable in a real election. Transaction systems are not exactly new and neither are RAID arrays to prevent losing votes in the case of disk failure.

The only trick (other than a smooth UI) is to get the user program to send the votes to a central location. The must have been a thousand programmers in Brisbane alone who would have had the skill to do that.

You send them on a portable medium in a truck/car/whatever with a cryptographic system to make sure people can't modify them in transit or replace them in transit.


These systems aren't rocket science, they're student projects. If I had to do it again, I'd implement the whole thing in Java with a SQL backend. The java could be compiled on a single system, and then downloaded by the client voting systems on startup. Thus the police only need to audit one machine. With a team of 10 people, the whole thing could be designed, implemented, tested and documented in 6 months. If you add in an engineering team to make beautiful custom boxes (running *NIX), with nothing but a monitor, ethernet port and power switch, it could be shipped as one purpose built product.

It has in fact been done in less than seven months. http://www.softimp.com.au/evacs.html, but without using idiotic buzz word crap and instead suitable tools.

Google turns up one result for an effort in Australia, but I can't find any cost for the system, or a download link. There's also this post about a Python project, which appears to be talking about The Open Voting Consortium, which has a SourceForge project page.

So it seems that there are movements happening, but these don't seem to be getting anywhere quickly. Does anyone know of any other projects?

The whole system is open to public scrutiny - several people have reported bugs, including an academic. Nice contrast to the DMCA...

Better yet, since members of the legislative assembly of the Australian Capital Territory are chosen using Hare-Clark, the ACT electoral commssion has contracted a company to build free software electronic voting and counting systems to conduct elections using the method.

PR systems like Hare-Clark are somewhat difficult to count by hand, and the most accurate algorithms, such as Meek's Method must be done by computer.

Incidentally, complexity is one of the major problems facing adoption of proportional representation schemes... the mechanics are somewhat difficult to explain to nontechnical voters, and thus debates on the issue lend themselves easily to spin and misrepresentation.

(The other major issue, of course, being that PR tends to threaten established politicians and other elites... here is an interesting site, for instance, that discusses the impact of PR in New York after communists and blacks were elected using the method back in the 1940s).

Better yet, since members of the legislative assembly of the Australian Capital Territory are chosen using Hare-Clark, the ACT electoral commssion has contracted a company to build free software electronic voting and counting systems to conduct elections using the method.

PR systems like Hare-Clark are somewhat difficult to count by hand, and the most accurate algorithms, such as Meek's Method must be done by computer.

Incidentally, complexity is one of the major problems facing adoption of proportional representation schemes... the mechanics are somewhat difficult to explain to nontechnical voters, and thus debates on the issue lend themselves easily to spin and misrepresentation.

(The other major issue, of course, being that PR tends to threaten established politicians and other elites... here is an interesting site, for instance, that discusses the impact of PR in New York after communists and blacks were elected using the method back in the 1940s).

Is there some kind of financial relationship between Wired's owners (Conde Nast publications) and the Australian voting company?

For the record, No.

BTW Software Improvements isn't a "voting company", we do stuff like Satellite Avionics etc. Like the software on FedSat, which recently rode out the biggest solar storm in recorded history. As you can see live on the web.

<sarcasm>Of course our greatest achievement is our Time Machine.</sarcasm> eVACS was first used in 2000, before the Diebold controversy erupted. So much for us trying to cash in on it.

Finally, if you think the system's out to bamboozle people - why not tell us exactly how? The Operating System, the Compiler, and the Code are all Open Source and available for all to criticise. An independant lab "fact-checked our asses", you can too. In fact, if you read some previous /. posts, you've been able to for some time.

Finally, a disclaimer. I work for Software Improvements, along with Matt Quinn. He's an unregenerate Leftie, I'm a Right-Wing Neocon Deathbeast. We keep each other honest.

Or why not look to an open-source solution available under the GPLd that has already been used in parliamentary elections in Australia.

1. Open Source Code
2. Open Source OS
3. Open Source Compiler
4. Standard PC Hardware
5. Independantly Verified by both Electoral Authorities and Independant Labs
6. In 12 languages
7. Audio help for vision-impaired voters
8. And actually used in 2001 government elections

1. Open Source Operating System
2. Open Source Compiler
3. Open Source Code

That's Australian management all over.

But not every company is like this. OK, I'll name names: I work for this mob. The CEO gets less than the senior technical people. Every 2 weeks, all employees get together and we discuss the firm's accounts, what our plans are, how we can help each other. It's a (literal) Soviet. And we do all sorts of interesting stuff, like spaceflight avionics, help people get CMM 3+ etc. We work 9-to-5 in theory, but closer to 10-6 in practice. We have lives, and consider long hours to be a sign that we've screwed up in resource planning. It happens, but rarely.

Dress Code? "Whatever's appropriate". When flogging a tool to IBM, suit, tie, etc. Otherwise "civilised", whatever you're comfortable with. Some wear suits by preference, some not.

No, I wouldn't leave this place for 10x the money I'm getting. After 22 years in the business, I know how lucky I am.

Damn, where are my mod points when I need em.. I'd give a "+1 Extremely Useful" to this.

For my sins, I've had to re-vamp our company's website. After a lot of experimentation, here's some design principles:

• Sadly, CSS has too many different interpretations of what it should be doing on different browsers. This is a Royal Pain.
• Take Accessibility by the blind seriously. If doing work for the gummint, this is mandatory. In any event, it's unethical not to do so.
• Javascript is right out. Also a royal pain. This is because of security issues, a growing number of corporate clients have Javascript permanently turned off. This means though that you can't do effective versioning.
• ASP and other MS-only stuff - don't even think about it, unless you want to appear like a total dork. Your next employer might (will) want to see what you've done, and if it's nothing but FrontPage-wizard generated MSTML, then you'll be shown the door, unless they're an MS shop.

Damn, where are my mod points when I need em.. I'd give a "+1 Extremely Useful" to this.

For my sins, I've had to re-vamp our company's website. After a lot of experimentation, here's some design principles:

• Sadly, CSS has too many different interpretations of what it should be doing on different browsers. This is a Royal Pain.
• Take Accessibility by the blind seriously. If doing work for the gummint, this is mandatory. In any event, it's unethical not to do so.
• Javascript is right out. Also a royal pain. This is because of security issues, a growing number of corporate clients have Javascript permanently turned off. This means though that you can't do effective versioning.
• ASP and other MS-only stuff - don't even think about it, unless you want to appear like a total dork. Your next employer might (will) want to see what you've done, and if it's nothing but FrontPage-wizard generated MSTML, then you'll be shown the door, unless they're an MS shop.

Some data about FedSat, a Scientific R&D microsatellite that's due for launch on the next H-2A from Tanegashima:

Firstly, the on-board software is in Ada-95, using the 3.13p version of GNAT as the compiler and RTEMS as the Run-time Kernel. Both Open-Source, and the 'p' in "3.13p" means public, free-as-in-beer. The on-board software was developed mostly by Software Improvements, a bunch of software professionals who are heavily into SLCMs, CMM, etc. And lo, it worketh, on-time, under-budget.

The ground station is another matter. OASIS was tried, but couldn't do all that was required. In a spectacular case of less-than-wonderful-judgement, VB on Windows was used for development. After a while, they got some software pros to work on that one too, rather than the hack'n'slash electronics engineers. Well, it partially works - enough so that a few months or years after launch, it will do most of what's wanted.

Moral: if it's important, and has to work first time, every time, do it in Ada and open source, and use the principles described in the parent article. If it's not so important, and can be fixed up afterwards, you're still better off using Ada, though Java's a good second choice. The only reason Ada's so good is that it makes it easy to adhere to good software engineering principles, such as teamwork, smart design, and open standards.

• Separate Public view and private view -> Interfaces between teams made easier
• High-level (generics, tasks etc) and low-level (address clauses) features ease design task
• Standardised - Ada-83 was standardised as Mil-Std-1815A in January 1983 and later as ISO/IEC 8652:1987, Ada-95 as ISO/IEC 8652:1995.Copies of the LRM (Language Reference Manual), style manual etc are all freely available on the web.

A.Brain, Rocket Scientist

A Fully Tested Open Source E-Voting GPL'd system is available on the web.

It was developed within 27 weeks for about $100,000 US. Multi-language, using standard COTS hardware and OS. (The compiler and OS had to be open-source too of course - Debian and gcc). It has been used in a state election in the Australian Capital Territory, the equivalent of the District of Columbia. There's an Executive Summary of how well it did, warts and all. A PDF of the full report is also available.

/. readers will be most interested in the technical description. Oh yes, the code's available as a Zip file here.

The whole point about e-voting software is that it has to be open-source. The hardware has to be available for inspection at any time too, along with the OS source and the compiler source as well. The situation as described in the original article has a strong piscine aroma.

Disclaimer I work for the mob that did the Aussie system - though I was busy making spaceflight avionics software rather than election software at the time, it was another team. They Did Good.

Circumstances alter cases is the old legal truism. It certainly applies here.

Examples from an IT career spanning over 20 years and several Continents.

I've worked in a company where I felt I was underpaid by market rates, got a job offer "out of the blue" at a much higher rate, went to my supervisor and said "I can't afford to keep working here.". He looked at the figures, said "You're right, and we can't afford to pay nearly that much, so Good Luck and do you need a reference. Your work here has been a great contribution, I don't know how we'll replace you, but we'll manage. You must act in your own best interests." I left with some regret, but recommended them as a good bunch to work for - and delayed as long as I could in order to do a complete handover.

I've worked for a firm where much the same thing happened, but I decided that the pay wasn't as important as how much I enjoyed working there. I helped finish off the project, then started looking for a new job with as much interest but higher pay. I got it - again, I was truly being underpaid.

I've had a job where they made an immediate counter-offer. Within a month they upped the salary of everyone else to match market rates, as they realised they'd been underpaying everyone. I left when they ran out of interesting work in my niche, but again, would recommend them.

I've had a job where they made an immediate counter-offer, but told me not to tell anyone else so they could keep underpaying them. I RAN from this bunch of sleazebags - I hadn't realised they were so - er - "ethically challenged".

I've worked for a company where the employees got together annually, we were given the financial data of the company, and we all decided what our pay should be for the forthcoming year. I might add that when times were lean, it was the more senior people who asked for their pay to be cut, as we could afford it more than the junior people - some of whom even got rises, as they deserved it, and we could afford it. And with our longer experience, we realised just what a rare and wonderful situation we were in. Do you know what it's like when a bunch of people you respect insist that you get a pay rise because of your contribution to the firm? The money amount is irrelevant, some things you just can't buy. (Actually, I'm still working for this firm.)

So Circumstances alter Cases. There are firms out there that have loyalty to their employees, and expect (though don't always receive) employee loyalty in return. There are firms that think the concept of "loyalty" in any direction is irrelevant, and settle for impersonal fairness. There are those that try to screw the employees for as much as they can - rather too many of these, alas. And there are firms which are the employees, with everything out in the open. Damn few of em, unfortunately.

One tip: if Money is the be-all and end-all when it comes to benefits, rather than job interest, professional development and career-growth, the greater the odds are that they're out to screw you. Look at the training budget. High means they invest in you, and will try to keep your loyalty. They'll pay market rates to everybody if they can, out of sheer self-interest. Think long and hard before jumping ship to someone paying more but with a lower training budget - you may well regret it.

A page that reports what your browser is telling it, and what the page thinks is actually the case, is here Yes, it will detect Opera, even if Opera's masquerading as something else.

I ripped off^H^H^H^H^H re-used the code from elsewhere - leaving attribution in the source, then modified it a bit. If anyone knows a better bit of javascript to do this, I'd be interested.

Any relative novice who aspires to the title of Webmaster could do worse than having a look at the whole About This Site section, which deals with making pages browser-agnostic, fast to download, accessible to the visually impaired, and not reliant on plug-ins or even scripts. I'm the author BTW, and most certainly not an expert, or even good. Just better than the Frontpage scriptkiddies that masquerade as 31337 htmlasters. Anyone who can give me some more tips on how to improve the site, feel free to contact me.

A page that reports what your browser is telling it, and what the page thinks is actually the case, is here Yes, it will detect Opera, even if Opera's masquerading as something else.

I ripped off^H^H^H^H^H re-used the code from elsewhere - leaving attribution in the source, then modified it a bit. If anyone knows a better bit of javascript to do this, I'd be interested.

Any relative novice who aspires to the title of Webmaster could do worse than having a look at the whole About This Site section, which deals with making pages browser-agnostic, fast to download, accessible to the visually impaired, and not reliant on plug-ins or even scripts. I'm the author BTW, and most certainly not an expert, or even good. Just better than the Frontpage scriptkiddies that masquerade as 31337 htmlasters. Anyone who can give me some more tips on how to improve the site, feel free to contact me.

A page that reports what your browser is telling it, and what the page thinks is actually the case, is here Yes, it will detect Opera, even if Opera's masquerading as something else.

I ripped off^H^H^H^H^H re-used the code from elsewhere - leaving attribution in the source, then modified it a bit. If anyone knows a better bit of javascript to do this, I'd be interested.

Any relative novice who aspires to the title of Webmaster could do worse than having a look at the whole About This Site section, which deals with making pages browser-agnostic, fast to download, accessible to the visually impaired, and not reliant on plug-ins or even scripts. I'm the author BTW, and most certainly not an expert, or even good. Just better than the Frontpage scriptkiddies that masquerade as 31337 htmlasters. Anyone who can give me some more tips on how to improve the site, feel free to contact me.