diebold admits voting machine misses votes
OK, this just on slashdot.org.
“Premier Election Solutions (a subsidiary of Diebold) has acknowledged a flaw that causes the systems to lose votes. It cannot be patched before the election and the machines are used in half of Ohio’s counties, but they are issuing guidelines for avoiding the problem that presumably contain a work-around. While Diebold initially blamed anti-virus software for the glitch, they have now discovered that the bug was their own fault for not recording votes to memory when the cards are uploaded in ‘certain circumstances’ — something their initial analysis missed. It would be nice to hope that Ohio poll workers would be tech-savvy enough to make this a non-issue, but they had poll worker shortages last year and might need tech-savvy people to volunteer.”
OK. So colour me stupid or something, but really, isn’t a VOTING MACHINE one of the most basic things which could possibly be written.
Lets take a few things out – security for a start. Lets assume, for a moment, that it’s in a nicely sealed box with a large man standing next to it with a pitbull. How hard should it be to have a configurable list of candidates, and allow the user, who walks up, to enter some form of ID (eg a number), validate it, record their selection, and thank them.
Really, that’s computer science 101 stuff. In NZ, people do this with a pen and paper.
Fair enough that the security around it is more difficult, but with the likes of Steve Gibson, Peter Guttman and their international equivalents around, that part should be fairly easy to cover, so you need neither man nor dog to be present.
So really, how hard should it be to do this, for a well funded company. Given the press around it, it’s obviously harder than writing an operating system – and if a group of paid, loosely organised hackers can produce Linux, I’m sure a voting machine shouldn’t be too hard.
Really, it’s sad and stupid. And an obvious thing for the likes of Richard Stallman to get his teeth stuck into. Something normal people would actually care about – and where openness is a fantastically good idea.
The real problem with Diebold (according to many stories published, mainly back during the last US Pres election which Dubya “won”) is that they are untouchable in court and they are federally exempt from having their software audited. How can you guarantee accurate and impartial election results when the company responsible for collecting the results is free from regulatory oversight?
I would love to see someone like ‘rms’ develop an open and infallible voting machine (obviously assuming the hardware itself isn’t compromised) – beyond the US [which are probably doomed to repeat this process with the rather shady way they put contracts out for tender] this would obviously be of benefit to a lot of emerging nations etc. trying to remove some elements of corruption from elections.