A ruling came out of the Florida courts yesterday that’s managed to pique my interest a bit. In the case, a group of accused drunk drivers requested access to the program code for the breathalyzer that was used to document their blood alcohol levels; the court agreed with their request, and ordered the state to provide them with the code. The kicker is that the manufacturer of the breathalyzer claims the source code as a trade secret and is refusing to surrender it to the state, meaning that all of the drunk driving convictions obtained by using the device can now be called into question (and potentially overturned).
To me, this makes perfect sense. If a tool is going to be used to document some fact that’s used to make decisions about right and wrong — criminal and legal — then that tool better be as transparent as possible so that experts can be sure it works the way it’s advertised. In medicine, we would never make clinical decisions based on experimental or unverified test results; in fact, there’s an entire certification process through which new laboratory tests must be put before they can be used to make clinical decisions, and that process forces the people who develop and manufacture the tests to open their processes up to independent experts for verification. Why should the criminal justice system treat tools used to gather evidence in a different manner? (This is all the more important in the Florida case, as the breathalyzer in question has a questionable accuracy record (PDF), and was even subject to a recent software recall.) Conversely, why would a police department feel comfortable using a tool that operates in a completely hidden, unverifiable way?
It makes me happy when rigorous scientific standards find their way into places they logically belong.
Transparency? It’d be a nice thing for, oh I don’t know…voting machines? To take one obvious example.
• Posted by: TheBradYou make such petty demands, Brad, you really do…. ;)
• Posted by: Jason on Nov 4, 2005, 6:30 PMWhile I’m sympathetic, Jason, is there a difference between the vetting procedure and what the court’s doing here? I doubt that the medlabs have to divulge all their secrets to the open world, but they certainly are doing so to some experts. There’s a big gap between this and open code, I’d argue.
It’s not that I disagre with the courts, because I agree with them. However, opening procedures to a third party, whether an NGO or the government, is a far cry from forcing open a codebase.
• Posted by: Geof F. Morris on Nov 4, 2005, 9:34 PMGeof, read the article — nobody’s being compelled to release the code to the public at large. Instead, the court is telling the state (the only party to the suit over which the court has jurisdiction) that it must release the code to a single person (Harley Myler, an electrical engineer appointed as an expert by the court) for an independent review. The problem is that the state can’t comply with the ruling… because the company who manufactures the breathalyzer won’t release the code to the state in the first place.
So in the end, it isn’t about a company that wants to prevent the public from getting access to its code, but rather about a company that won’t even allow its code to go through an independent review to verify that it does what it claims to. So it seems perfectly logical to me that the court system can’t accept data generated from the code as verifiable (especially since the device run by the code has a record of inaccuracy!).
• Posted by: Jason on Nov 4, 2005, 9:44 PM[Dangit, I hate it when I adopt a Slashdot attitude.]
That’s cool. Independent reviews are always good … and I’ll definitely also concur with Brad’s voting machine comments, too. :D
• Posted by: Geof F. Morris on Nov 5, 2005, 4:11 PM