Thanks for the clarification. Obviously I misunderstood the court ruling. Either way, if the state does manage to squeeze the code out of CMI (assuming they intend to make any effort to do so), I think this guy is going to find this was all for naught.

Not to quibble, but . . . okay, to quibble:

The court didn’t say that the State has any right to that source code. The State has been saying that it doesn’t have any right to it and so can’t get it, and what the court did now say was that, if the State wants to rely on the machines in prosecutions, it will turn over the code. So, the State can either say “we can’t get the code”, and then try to make all of those expensive Intoxilyzers into Heathkit-inspired smoke alarms or wifi modems or really cool ashtrays, as long as they don’t try to cite to them in court anymore, or the State can say to CMI “no more sales until we get your code, plus we’re gonna sue you for what we’ve paid for all these machines.”

To what lengths WON’T you go in your quest for trains? Good grief, drunk driving is an evil brought to us by sprawl?

(Don’t forget hemorrhoids from all the driving, spousal abuse from lawn hate, and racial divide due to sidewalks.)

“Bobby, perhaps you’d care to enlighten this software engineer to the differences between code that runs in a programmable controller vs code that runs on a computer?“

Uh, sure, but could you move over a bit first? You’re dripping your condescension down the front of your new Help Desk tee-shirt and I just washed up.

The difference is in the relative simplicity and brevity, as you (I suspect) already know. You have very few functions being substantively run by code. The actual measurement of alcohol is done by an infrared spectrophotometer, which is giving its own variable readout directly and which is not subject to further “interpretation.” You have 3 basic cycles that can run, and the programming involved essentially only sequences the individual cycle types, (of “breath”, “clean air sample”, or calibrate”), and allows the user to select the cycle pattern (A-B-A, or A-B-A-B-A, for the states that require double-testing). The calibration only involves testing a known vapor sample, and the air blank cycle is just a test on clean air, so even the three cycles are designed, in the logic portion that directs them, pretty much the same. the other functions mostly involve a simple internal database of tests performed and calibrations checked.

Keep in mind that I was responding to Mr. anti-Diebold, whoo’d seemingly like us all to fervently believe that The Man has taken control of our TVs and our voting machines and our Intoxilyzers and probably airplanes as they pass World Trade Centers. I’ll stand by my comment about his commnt (which, to remind you, was “in the cases where source code was able to be reviewed they saw horrible coding standards what would not be acceptable in any modern business or government agency.“)

At first glance the source code cases look like they could be more about imposing a lot of unpalatable costs and hurdles on the prosecution than they are about the actual accuracy or inaccuracy of the machine. The defendants aren’t subpoenaing the information themselves from the machine’s maker. They’re demanding the State get it. This is how lawsuits sometimes turn from searches for the truth into battles of the pocketbooks. I suspect the Department of Public Safety is mostly interested avoiding lengthy hearings on the science and computer codes every time they try to pull a license for DWI. Thier chief claim in the Minnesota Supreme Court was that once an instrument is certified by state studies and tests, Minnesota law prevents relitigating that science in every case.

The Minnesota Supreme Court didn’t decide the issue of what the State has to provide. It just decided not to interfere before trial with a district court order for the State to disclose the code. Another district court judge could look at these same knotty issues and rule the opposite way. The State owns some code for the machine that is specific to Minnesota but claims it does not own the “complete” source code demanded and claims it is prohibited by copyright from disclosing it. The Minnesota Supreme Court punted on this point. “Having carefully reviewed the record presented and the arguments of the parties, we conclude that we cannot decide the copyright issues raised.”

you mean you use a different compiler? wtf?

Bobby, perhaps you’d care to enlighten this software engineer to the differences between code that runs in a programmable controller vs code that runs on a computer?

I don’t agree necessarily that the code for breathalyzers (or voting machines) needs to be completely open source, but review of each version of the code by multiple independent panels of disinterested experts before being trusted by courts is a must. Perhaps this source code was already reviewed in such a manner for all I know. I’m not sure what the protocols are, if any.