Slashdot Mirror
Breathalyzer Source Code Revealed
Nonillion writes "New Jersey attorney Evan M. Levow was finally able to get an order from the Supreme Court of New Jersey forcing the manufacturer of the popular Draeger AlcoTest 7110 to reveal the source code. Levow turned the code over to experts, Base One Technologies, to analyze. Initially, Base One found that, contrary to Draeger's protestations that the code was proprietary, the code consisted mostly of general algorithms: 'That is, the code is not really unique or proprietary.' In other words, the 'trade secrets' claim which manufacturers were hiding behind was completely without merit." Following up an earlier discussion here, the state of Minnesota has (without explanation) missed a deadline to turn over the code for a different breathalyzer.
My guess is that 99% of proprietary code contains a big trade secret: The secret of just how crappy the source code really is.
If they were expecting their code to be opened to the public, they would have taken the effort to fix up "spaghetticode.inc" which contains the single comment "//This works though i'm not sure why... clean up l8r!!!!".
It's a device intended to nab as many people as possible. The more people it "saves" from being killed by drunk driving the better. Accuracy doesn't matter, legal limits don't matter. ZOMG ALCOHOL!!! = Jail. Fines. Moral superiority. If police departments actually intended to serve the public, they'd come up with a more reliable system subject to completely public scrutiny and be glad to instill public trust in their methods by doing so.
Flip it to another tool used for criminal convictions: if DNA were a public, proprietary process through only two or three companies nationwide and they refused to show anyone how it worked, would you trust them? Absolutely not.
The unfortunate reality that the laws are trying to deal with, though, is that it is essentially impossible for law enforcement to spot all the drunks on the road, and deliver you home as you suggest (imagine the logistics of that: you could put all the police in this country on that duty full time!, and still not have enough cops).
Worse, it won't even be near to possible for them to indentify all of the sufficiently impaired so as to protect the rest of us from their idiocy.
What drunk driving laws do is create an incentive for everyone to voluntarily police themselves, and to act more responsibly. If you know you run a risk of a long incarceration just for drunk driving, you may not take my life into your hands by getting behind the wheel and driving the same roads as I do. If you (or most of these drunk idiots) know that the only penalty for getting caught is being taken home, then you'll be much more encouraged to just take your chances with my life, rather than deal with the inconvenience and cost of a taxi ride.
Drunk driving laws disencentivize behaviors on an individual basis that normally have unfortunate incentives on an individual basis, but have an extremely high average cost for the rest of society. This is also why no-sleepy-driving and no-cellphone-driving laws are a similarly good idea.
"Who is the Journal of Quantum Physics going to believe?" --Stephen Hawking
Here's a better idea: don't drink and drive.
True confidence comes not from realising you are as good as your peers, but that your peers are as bad as you are.
I think it's great the source code is getting out, and that we'll find out which devices are crappy and which are better.
But in the end, I don't think any of this matters. Drunk drivers are not prosecuted based on roadside breathalyzer tests. They are prosecuted based on tests done back at the police station using either a blood test or a much better lab-quality breathalyzer. These instruments are regularly tested in a way that makes it easy to convince a jury of the validity of the results. I've seen some of the corresponding tests on a roadside breathalyzer, and they convinced me not to trust the device.
So, it's good advice to decline the roadside tests.
I have an even more important question: Does the friggin' device work? I agree that reading through the observations, the code doesn't instill confidence. But the real important question is whether or not it works. There must be some requirement as to how many false positives/negatives are allowed because no matter how good your code, nothing is infallible. So what is the requirement in terms of acceptable false positives and/or false negatives, and does the device meet that requirement?
Is there is a real and legitimate belief that this device doesn't work? Or is this just some escapade launched by an attorney to free a guilty drunk driver?
Yes, but how many karma points are *gained* by drunken slashdot posting? I'd say it's probably an order of magnitude greater than the points lost =]
Everything I need to know I learned by killing smart people and eating their brains.
If the law says 0.05 is fine then we should hold ourselves to it. If we wanted 0.00 then write the law as such. Why is carrying out the law properly even an issue?
Considering this is shipping code in a device that doesn't exactly do automatic updates over a wireless network, I'm not sure when, exactly, you're anticipating that this testing" code will be replaced with the "real thing". You'll forgive me for thinking you're taking a fairly blase attitude towards the obviously complete lack of coding standards, formal oversight, or rigorous vetting in code that can quite literally destroy someone's entire life based on the output.
Playing around with temporary hacks is fun for a shareware app, not an officially-sanctioned law enforcement device that decides whether you were the victim of an accident or the perpetrator of a felony.
Recursive: Adj. See Recursive.