Microsoft, Unisys & Dell To Make New Voting System

About twenty million - alright, slight exaggeration, but a whole bunch of people sent the story about Microsoft, Dell and Unisys to build a new voting system. Microsoft will do the software, Dell the hardware, and Unisys will assemble the systems.

Comments

MIT, Caltech also developing a voting machine.

[Teratogen]

MIT and Caltech have already announced plans
to develop a voting machine:

http://web.mit.edu/newsoffice/nr/2000/voting.htm l

Personally, I trust MIT and Caltech much more
than I trust Microsoft and Unisys.

Re:MIT, Caltech also developing a voting machine.

[ElJefe]

Here's the corresponding Caltech link:
http://www.caltech.edu/events/mitcit/citmit.html

-Chris

Re:Schneier on voting systems

[Seumas]

What I don't understand is why the same government that tried to tear Microsoft apart because of their unethical business practices would want to then employ them in the pursuit of a new voting system -- something which one should aspire to include only those participants who are extremely ethical.

Remember Netscape Engineers are weinees and 'leasing the operating system' and '.NET' and recently, the backdoor account in Interbase (granted, not MS but just an example of the type of things one could not put past Microsoft)? -- Is this really the kind of risk we need to put our voting system in?

The whole idea of a Windows-based system is frightening. Why not go for something that at least will allow an uptime long enough to complete the full voting day? Maybe Solaris or Linux or a reduced size, limited function distrobution of their own?

Further, these are obviously going to be networked throughout the country, to a central city computer, then to a central state computer and then to a central national computer (sure, you could go directly to a national machine, but I think there may be some federal/state issues that would make regionalizing each function of the process preferrable). As we all know, anything that is on a network is potentially crackable.

If anyone needs proof of that, look at the NSA, CIA, FBI, IRS and other government websites which have been cracked in the past.
---
seumas.com

"Microsoft will do the software..."

[radiashun]

do I have to even try to poke fun at that?

Re:"Microsoft will do the software..."

[Silverlock]

Does this mean that if we have Back Orifice, we get as many votes as we want?

[click][click][click]...

"What are you doing?"

"Making damn sure Jello Biafra wins this one."

Silverlock

Re:"Microsoft will do the software..."

[nomadic]

Just click on the "Election Fraud Wizard" for easy setup...
--

Re:"Microsoft will do the software..."

[tonyPick]

"You have registered your vote, please restart the electoral system for this to take effect"

Re:"Microsoft will do the software..."

[me.at.work]

"It looks like you are trying to rig the election, do you need help?"

Re:"Microsoft will do the software..."

[fortunetroll]

Microsoft is well known for producing top of the line quality software and is well respected around the industry as well as unanimously praised in all forums as being the saviours of mankind and the leaders of the free world.

You better start believing it, boy... they run the elections.

On Monday mornings I am dedicated to the proposition that all men are created jerks. -- H. Allen Smith, "Let the Crabgrass Grow"

3 of a kind

[Xtacy]

heh thats all we need, ms to screw it up, dell to charge a fortune for the screwup and unisys to patent the screwup after everyone has been using it for many years

Re:Once, just once...

[warpeightbot]

There is no reason to re-invent the wheel. The Brits (of all people) have done it for us. They wrote a program called FREE and published it under (what else) the GPL. It's a java-based secure client server that will run on most popular OS's (W32, Linux, Mac, OS/2!) and (get this) is *already in production*.

I think we have to insist that any voting system implemented be Open Source, and specifically GPL, so no one can go and put backdoors in it without someone being able to find them. Furthermore, I think we can make it stick, to-wit: Proprietary software must be paid for, right? Which would make it an unfunded mandate, and thus unconstitutional.

Folks, get on the phone to your congresscritter, write letters, whatever, but we can, and have to, stop this right now before it spreads. I'm not going to stand for spending several million dollars of MY MONEY to develop something that has already been developed and is out there for the asking... something I will never trust unless it is open source, and neither should anyone else.

Once more into the breach, dear friends, and we can stop Gates' final attempt to take over the world, and have it for ourselves, and set it (ahem) FREE.

--
If you want to end war and stuff, you've gotta sing LOUD.
-Arlo Guthrie, "Alice's Restaurant"

Bah!

[Deluge]

C'mon, really. All this phenomenal expense, and for what? For a system which will have a million bugs (and it would anyway, even if MS wasn't involved, so shush) and which people will be apprehensive about using and which will cause an even BIGGER mess in case of a screwup, since you can't just manually examine an electronic vote.

It's been said in almost every one of the voting stories thus far - why not go with a system like Canada's? Simple X in a box, hand counted, done in a few hours, no ambiguities, no problems. Ugh.

---

Re:Bah!

[NumberSyx]

The problem with electronic voting is there will be no way to verify a vote after the fact. The software will be by Microsoft and therefore proprietary. The Hardware by Dell and will therefore be proprietary. This will be cross checked by a "Third Party" Unisys, but of course there will be no means to independently check for backdoors or vote skewing code, we will have to take their word for it. Even if everything is on the up and up to start with, will this system be on a network ? If so, we all know how safe and secure networks are. Can anyone say "Self Propagating VB Script".

I say No, lets go back to the old fashioned way, a printed ballot with check boxes and a pen. Then each precinct is responsible for counting their votes on location and accessible to the public. Once the count is done, the votes are sealed, locked up and impounded to prevent tampering in case a recount is necessary. Results are then reported to the county, first by phone, then later in writing, which then reports to the state, who then reports to the Federal Voting board. Everything is nice and clean, with a clear paper trail of accountability.

Jesus died for sombodies sins, but not mine.

Re:Bah!

[Coulson]

This could go a long way towards reducing voting confusion (ala butterfly ballots). You could have a well-layed-out touch screen rather than being limited to the size of a paper ballot. Voting in an elderly county which has a lot of with people poor eyesight? Increase the font size.

At the end of the session, after people have picked their candidates, you can present the information to them again and ask them to verify that the choices are correct. No dimpled chads, no half-punched holes, no double votes. It won't eliminate problems, mistakes, or complaints entirely, but it allows options which are unavailable on paper.

So the system may be buggy at first: treat it as a critical systems project (life support, chemotherapy machines). The best part of all -- instant and accurate tallies, where the numbers stay the same no matter how many times you add up the votes!

Re:Bah!

[FFFish]

To elucidate on this point a bit further, Canada's system *works.* We might not like the results (and who does? Inevitably, a politician is elected. Seems a rather unfair consequence, really.), but it works.

Voting centres are set up most anywhere that there's adequate floor space: generally, gymnasiums and halls. A greeter asks to see your voter registration card, and then directs you to the appropriate tables; a wholly unnecessary step, because the tables are clearly marked with a pair of letters that indicate what range of names ("Aa"rdvark to "Bo"gart, etc) they're taking.

Being Canadian, you line up nice and neatly, and patiently await your turn to vote. Pushing ahead in the line, or making catcalls at a particularly slow voter, would be un-Canadian, and we'd all have to scowl at you and possibly mutter under our breath.

Once you get to your voting table, you're greeted by at least two, and perhaps three, volunteers. They're from opposing parties, to keep each other honest.

One of them takes your voter registration card and scratches your name from the master list. The other waits until that process is complete, and then tears a voter card from a booklet. You're then instructed, using the same words you heard given the previous voter, on how to clearly and properly mark the card. The volunteer pre-folds the card and hands it to you.

A short cardboard booth is set up at the end of the table. You can see over it, but no one can actually see what you're marking down. It's a little discomfiting; seems to me that last time, our booths completely hid us from sight.

The voting ballot has a black background. 1.25cm (that's half-inch, in obsolete terms) white strips line the page. In each strip, printed large, is the name of the candidate and their party affiliation. Directly beside the white strip, to the right, is a white circle.

The names are in alphabetical order, last name first, first name last. You place a mark across from the candidate you want to elect. Because each region elects only a single Member of Parliment, you only mark off one circle.

You fold the card, and fold over the retaining flap, so that the card doesn't flop over. You hand the card to the volunteer, who makes sure that the flap is secure, and then drops it into the vote box as you watch. I believe there's every chance that the volunteer asks if you marked off one, and only one, candidate.

And away you go, happy to have participated in a futile ceremony that will surely see no real changes made to the social, political or economic fabric of the country. No, I'm not bitter. Not at all.

After the polls close, the volunteers dump the votes out on the table and begin counting them. There's a paid overseer with a big bullwhip that makes sure they do the job quickly and correctly. Quite possibly, there are plenty of party representatives watching over the vote-counting process.

There are no pregnant, well-hung, dimpled chads. There's either a clear mark in one circle, or there's an invalid ballot. The count goes quickly. All the ballots go into a lockbox, for safeguarding.

I suspect that only the Australians have a better system, and that only because it seems that their elections office is self-supporting, because it does such a fine job that it contracts itself out to provincial, municipal, union and other votes.

The American system, on the other hand, is appallingly asinine.

--

Re:Bah!

[Malc]

It has nothing to do with size. The Canadian system scales very well to Britain with twice the population. They too normally have the counts finished in one evening/night. It's because the counting occurs in parallel. It's the old divide and conquer approach.

Re:Bah!

[FFFish]

The residents of Nunavut voted on the name "Bob," but the government told them that that name wasn't allowed. So, yes, now they're Nunavut, and, yes, they are "created."

It seems to me that the USA has an orderly system of schools, despite having ten times the population of Canada. Is there any reason the school gymnasium couldn't be used as a polling station for the immediate area?

You see, the key to success hasn't anything to do with size: it's to do with having the polling stations a reasonable size.

My town's main polling station had about a dozen tables set up. Even if *every* person in this town--including children--were to have voted at this station (but we had three), each table would have processed only 2500 people during the day. It'd take well under and hour-and-a-half to tally those votes.

The hand-counted ballots aren't a problem: the USA can do that.

The problem is with a godawful ballot design, voter registration irregularities, voters being hassled by The Man while heading to the polling station, ballots lockboxes being lost, etcetera, etcetera.

Fixing the ballots would be a first, and probably small, step in fixing the tragicomically broken US election system.

--

Re:Bah!

[TrevorB]

Speaking of space...

The following was posted frequently to sci.space for the past 10+ years. I'm uncertain of it's origin. Perhaps it's relevant here...

During the heat of the space race in the 1960's, the U.S. National Aeronautics and Space Administration decided it needed a ball point pen to write in the zero gravity confines of its space capsules. After considerable research and development, the Astronaut Pen was developed at a cost of about $1 million U.S. The pen worked and also enjoyed some modest success as a novelty item back here on Earth.

The Soviet Union, faced with the same problem, used a pencil.

Online Voting...

[enneff]

I hope that when they try the online voting thing, they actually do it _right_. (a worthy goal for Microsoft)

The most obvious problems with online voting are identification and security. Voting in the United States is done via secret ballot, yet voters must first be identified. On the Web, once a visitor is identified, it is possible to track their movements and choices throughout the entire session.

Rather than pretending to not track this data, we should allow Web servers to behave as they already do and mask the identity of the user. PIN codes or site passwords, a la Amazon, are not a viable option as they are often written down and thus easily stolen, forgotten, or shared.

The best identification option may be the already present state driver's license or ID card. Many state cards currently have a magnetic strip that could be used to hold a voter registration ID. A better storage solution would be for states to use smart cards, like the new Visa and Amex Blue, as the basis of driver's licenses and IDs. The embedded chip could hold a variety of information, including e-commerce information, and would be read/writable. Along with helping the smart card industry gain about 250 million customers, the government will also need an equal number of smart card readers. These devices should be able to attach to computers, PDAs, cellphones, and Internet appliances.

The question is, can MS and co provide anywhere near this ideal?

Re:My letter

[BeanThere]

I still don't see how it could work, free or not. Even if you could get "the source", how would you know that that was exactly the same source code that was used to compile the system that will actually be used? It would be exceptionally easy for some corrupt person involved in generating the binaries to slip in a trojan before the software system is installed all over the country.

This problem would exist whether the software is (GPL-style) free or not, and I can't see how this type of problem can be avoided. Making it opensource would probably make it harder to find exploits (and you can be pretty sure that somebody somewhere *will* find an exploit) but you can't get around the basic problem of corruption.

Hand-counted votes are also open to corruption, of course, but on a much smaller scale - most people would only be able to affect the vote count in, say, their own voting station. But with a full-blown electronic system, somebody somewhere would have the potential to rig the vote on a massive scale.

Here's how this SHOULD work:

[PyroMosh]

Okay, I know I'm going to rile up a lot of people here, but I'm truly not flaming, trolling or whatever. I realy do feel that this is the "best" solution.

This past election was the second election I voted in and the first presidential election. I voted in Ocean County, NJ.

They use computers there.

They don't run windows. They don't run Linux. They don't run BSD, or any other general computer OS. They are dedicated voting machines with dedicated hardware and software built in to the microcontroler. Much like your VCR is a computer, but it only does one thing. Your Microwave oven has a computer in it, but it doesn't run Linux or a MS OS either.

The layout they used on this machine was simple. A bunch of buttons (similar to the kind under the plastic panel on your microwave, not the kind of buttons on your keyboard) under a paper overlay under a protective clear plastic sheet. There were probably hundreds of buttons, but only a few are ever used at a time, depending on the requirements of the individual election. Oh yes and there are LED lights next to each button, which holes in the paper so the proper LEDs could shine through.. There was also a keyboard (laid out alphabeticly) to enter write-in votes. This whole setup was perhaps 3'x3' or 4'x4' and took up the front "wall" of the voting booth.
I don't know how the machine OUTPUTS votes, (hard drive, print out, network, etc.) but everything I saw looked fairly straight forward, simple and easy to me. I didn't hear about anyone getting confused in my community. there were PLENTY of elderly voters and I was standing in line at the polling station for a while. The only confusion was over what lines to stand in (there were seperate lines to check your voter ID card and lines for the voting machine, in addition to splitting up lines by first letters of last names for check-in.)

Now, here's my take on how something like this should be designed:

The modularity of the paper overlay is a good idea and can be retained. Either that, or use expensive touch screens. Either is fine, and if cost is not an issue, I believe that touch screens would PROBABLY be a better plan since paper can (concievably) shift or slide. The important part here is that the layout (since either can be dynamic from election to election) needs to be CARFULY considered for each election.

The OS:
- Should NOT be Linux.
- Should NOT be BSD.
- Should NOT be Solaris
- Should not be Windows.
It should NOT HAVE an "OS" in the traditional sence of the term at all! It should have a simple "dumb interface" like a VCR, digital watch, phone, microwave, etc. When is the last time your Microwave "crashed"? If it's EVER happened, I'm sure it's happened less to you than any general computing OS you've ever used. Linux has crashed on me, Windows has crashed on me, Macs have crashed on me. My VCR? Hardware has failed (motors) but the software end of it has to my knowlege never failed.

Also, as to open source... this is a more touchy subject. I do agree that the code (probably C. the Elevator Principal applies perfectly to this situation.) does need to be audited make the source redily available. But I don't see any particular reason that it can't be developed by a private party / company/ whatever.

Several other people brought up concerns about "What if" people obtained the source code, edited it to their liking and installed it on the voting machines. Well, this isn't a problem with dedicated hardware/software on a microcontroler. When is the last time you heard of someone changine the software on their VCR? And without opening the thing up and breaking out the solder gun? And with people (Election officials) that don't want them to do this standing around watching to make sure they don't do this? In any concievable situation where this would be insecure, ANY method (pencil and paper for example) would be insecure due to the amount of corruption it would require. If anything, this might make coruption a little more dificult to pull off since it would require someone with in depth technical know-how AND would almost certainly take a conspiricy rather than a lone nut to rig votes.

Now for the other end of the equation. I believe (due to the companies mentioned) that they want to use a PC type of archetecture. I don't see why. It's insecure, unstable, and too generalized for the task at hand. Life support machines don't run Windows. Missle Guidance Systems don't run BSD. Power Plants don't Run MacOS. Why should this?

Now, I understand that this being /. and all, that one is expected to bash Microsoft.

I think this is lame.

Do I like Microsoft? Not particularly. Do I think they are evil? Only their business practices. But their software is the best thing out there for the home user. For the software I want to run, they and Apple are the only game in town. My web sites, however all run under Red Hat. I wouldn't DREAM of running any kind of site that recieved decent amounts of traffic under NT. However, why is Linux a good choice for VOTING MACHINES? Pick the right tool for the job people. PCs in general simply aren't cut out to going something like this. There's a reason it's called "general computing" because these machines have to be Jacks of all trades. The trade off is that they don't realy master any of those trades. They crash, they're often slower than a dedicated machine for the same task, etc.

Computers are not nessisarily bad for voting. In fact I encourage the use of computers. However, don't use general computers. Don't do this half assed. Don't try to shoe horn in the wrong tool for the job. Use a dedicated hardware/software solution.

Once, just once...

[ca1v1n]

Will they show us the freaking source? I think the point was exemplified by the battleship dead in the water, but I fear it may not have been taken to heart. Voting should be a completely transparent process with the sole exception of what goes on in the booth. I don't trust MicroSoft to even count right at this point. I want to make damn sure that nobody is going to work out a buffer overrun and move a few votes here and a few votes there and rig a close one. This needs to audited rigorously, and the source sounds like something that ought to be available at the least under the FOIA, and if the government can't get that right from MicroSoft, we shouldn't be using the software.

How to make electronic balloting work...

[TrevorB]

OK, a co-worker of mine were talking on the way back from lunch about a month ago on how to make an "electronic voting system" work. Let me first say that I'm Canadian and as FFFish's post states, our hand counted, hand marked ballots scale very well for 30 million, and I see no reason why it wouldn't scale well for 300 million. Whatever. Here's our idea.

The problems with a pure electronic system come with recounting, either due to a close vote, or to questions of tampering, is that there's no true way to do a physical recount. But regular ballots take "too long" to count.

Why not create a hybrid system. Each person comes to their polling station, is checked off a list and is given a "voting card", perhaps like a punchcard, but with no holes. They go back to the voting machine, close the door (or whatever), and insert their blank card into the machine. they vote via touch screen, and the result is printed on the card, which they (perhaps only as a symbolic gesture), insert their card into a box. When the elections end, you use the computers to tally the votes, but have the paper votes available if you need a true recount.

Now of course, all of this is meaningless unless the Americans can standardize their voting procedure. If one county has electronic voting and another punch card or another X'ed by hand, you're back in the same boat.

One thing FFFish may be wrong on is that it "Can't happen in Canada". Sure it can. Effectively, Bush and Gore tied within counting error. The same could happen in our system. Fortunately minority "tied" governments could exist in Canada without causing constitutional gridlock... (Just parlimentary gridlock, after a few weeks of which the government would collapse and we'd get to hold another election! :)

At any rate, I hope you guys can sort this all out by 2004.

Schneier on voting systems

[tarka69]

You might want to have a look at the Bruce Schneier (inventor of the Blowfish algo. and crypto pundit) on electronic voting systems.

Basically, he says they are a dangerous thing ...

The comforts you demanded are now mandatory -- Jello Biafra
--

Re:Yes!

[Black Parrot]

> I'll show George W. Bush that he's not the only one who can rig an election.

Be the first city council in your state to buy the all new Microsoft George voting machine! Features include:

• Microsoft's famous quality control and Dell's famous low prices.
• Rigorously tested by Mindcraft, Inc.
• Tasteful blue screen provides privacy between voters.
• Uses a special release of Borland's InterBase db server to keep track of the votes.
• Auxiliary USB port with no documented function.
• Coin operated version available by special order.
• Framed Certificate of Authentication, signed by Katherine Harris.
• Recommended by Jeb Bush and the RNC!

Testimonials:

• I'm not worried about fair elections anymore. -- George Walker Bush
• This simple device convinced me that the American Way (tm) would be to hold fundraisers at US corporations like Microsoft, rather than Buddhist temples. -- Al Gore
• I'd trade my entire cell for such a voting system! -- Slobodan Milosevic
• In jurisdictions where it has been adopted, I find that I don't have to waste scarce money by contributing to both sides anymore. -- Bill Gates

--

Someone has to say this....

[Stephen VanDahm]

OK, everyone knows that the only things that Microsoft makes well are joysticks and mice. So they should be doing the hardware. Dell ships computers with Linux preinstalled, so maybe they should be doing the software instead. As for Unisys, people who say "Democracy" and "Unisys" in the same sentence should be shot. Wait, did I just...damn....


========
Stephen C. VanDahm

Yes!

[Shoeboy]

This is so cool!
I need to go refresh my collection of rootkits.
I'll show George W. Bush that he's not the only one who can rig an election.
--Shoeboy

My letter

[Hard_Code]

Well, here's my letter...before I fire it off, anything obviously wrong or stupid?

Dear sir,

I am a software engineer employed in New York state. This morning, I
read on Reuters a report that Unisys Corp., Dell Computer Corp., and
Microsoft Corp., are teaming up to develop an electronic voting system.
Dell Computer Corp. is the number two PC manufacturer in the United States,
and Microsoft produces the de facto standard computer operating system
software, "Windows", used worldwide, and is currently in hot water with the
DOJ. I suggest that proprietary computer and software companies which
have large shares, if not monopolies, in their sector, may not be the
best entities to entrust with inventing a secure, fair, open, and
accountable electronic voting system.

In my opinion, both professional and personal, such a system is best
developed under a Free Software/Open Source model. I refer you to
descriptions of Free Software,
http://www.gnu.org/philosophy/free-sw.html, and Open Source,
http://www.opensource.org/osd.html.

Loosely, Free/Open-Source software is distributed under a license
which guarantees the freedom of people to obtain and inspect the source
code of such software. I believe only with this freedom of inspection
can we ever trust that software does what its originators claims it does.
Obviously this would be of utmost importance in an electronic voting
system. Furthermore, Free/Open-Source software has the additional
benefit of allowing many people to verify that the software is free of
bugs and performs as expected. This has the side effect of producing
software which is frequently more reliable and robust than software which
is developed behind proprietary curtains. Much of the infrastructure of
the internet is based on such Free/Open-Source software. Recently the
operating system Linux, also developed under this model, has been
receiving a good amount of attention and accolades.

Lastly I would like to point out that such a Free/Open-Source software
system for electronic voting already exists: FREE, "Free Referenda and
Elections Electronically": http://www.thecouch.org/free/. I am sure
there are other such Free/Open-Source projects.

As a professional, the practical benefits of Free/Open-Source software
are apparent to me. But as a citizen, I believe the social and ethical
benefits to state operated systems are paramount, and I could never
entirely trust with my vote, software whose source code I, or a trusted
party, cannot inspect. I kindly suggest that research into electronic
voting systems explore Free/Open-Source software possibilities.

IBM declined

[firewort]

IBM declined to get involved in voting mechanisms in the 50's under Thomas J Watson, and declined again this past year, when Lou V Gerstner was approached about the matter.

Voting is too important, and IBM has always chosen to keep their hands clean of ethical quandries, where possible. Besides, it's bad for business if fraud is committed on your machines.

What are the ramifications here? That if fraud is committed on a MS box, all /. will say, "yes, we knew it would happen, " the rest of the world will say "How could this happen?" and because MS is ubiquitous, it won't really hurt their bottom line, proving that they are a monopoly, regardless of what Judge Jackson's pursuit of them results in.

A host is a host from coast to coast, but no one uses a host that's close

Of course it can happen

[elroyjenkins]

If there is one single thing that I have learned about security is this... There is none.

If you want it safe, you dont want an electronic method like that. The more people involved, the more it (at least appears to) lessens the chance of seriously tainted results.

They shouldnt object to an open audit of the code, seeing how there isnt this huge demand for US Election Wizard 2.0, and they wouldnt have to worry about competitors.

I can honestly say that if I was involved in the project, and had access to the data in anyway, I would at the very least consider the idea of fixing the results.

Cute, but Incomplete Story

[llywrch]

Take a look at

http://snopes2.com/business/genius/spacepen.htm

Geoff