Train 49 - Why Late into CHI 7/29?

[neroden]

Amtrak doesn't need to subpoena the code. Proving what is or isn't in the code is a total waste of time. All they need to do is keep their eyes open and document the delays where their trains don't get priority. When the dispute is to be decided, the delay is what's important - not whether a real live dispatcher or an automated system caused it.

Well, yeah, but it's possible for the Class I to argue repeatedly that delaying a passenger train for a freight train was just an "innocent mistake" or "the best that could be done" if it's done by a live dispatcher based on undocumented "know-how".
If there's actual priority rules programmed into software which put freights ahead of passenger trains, that's a smoking gun. Those can be rather useful in proving one's case. For that matter, if the software clearly prioritizes passenger trains and the dispatcher overrode it to prioritize freight trains, that's also a smoking gun. And on the other hand, if the software clearly prioritizes passenger trains and the dispatcher followed it, then the Class I has an airtight defense. In short, if the rules have been put into code (including parameter tables), it's less of a he-said-she-said, and more of a "let's see who's telling the truth here".

 

[John Bobinyec]

That would be hilarious. I would love to attend a trial of some sort where Amtrak and the Class I were disputing what's in the code. The judge and or jury would be totally LOST. It would still turn out to be a he-said-she-said. Both sides would bring up their expert witnesses, all testifying for their side. How could a case ever be decided that way?

I wonder if the Supreme Court or any high-level court has had to wade through actual computer code yet. I imagine they will, someday.

jb

 

[neroden]

Sooner rather than later, since almost everything runs on code. That sort of trial is usually done with a special master. Fake "expert witnesses" who are talking bull are often thrown out under the Daubert standard by the special master before trial. (This is harder to do under the Frye standard, but can still be done.) Usually neither side wants a jury; even if they do, it's very easy for the judge to stipulate in the jury instructions that it is undisputed that the code gives priority to (whoever). Worth noting: judges are allowed to hire their *own* experts.

Microsoft faced the problem that the judges & juries it was faced with were perfectly capable of understanding what it had done. As a result, it deliberately obfusticated its code for the purpose of lying about whether different components were independent, something which came back to haunt them later (since it made the code unauditable and made it open to massive security holes). Microsoft lost their antitrust case anyway, partly because it was easy enough to prove that they'd deliberately inserted specific code for the purpose of sabotaging their rivals.

 

[jis]

This example has not much to do with the Microsoft case. This discussion is now GIGO.

 

[Green Maned Lion]

Why do we need to use arcane and asinine acronyms?

 

[John Bobinyec]

Why do we need to use arcane and asinine acronyms?

Because all the good ones are already taken?

jb

 

[Green Maned Lion]

Why do we need to use them at all? What the heck is GIGO?

 

[John Bobinyec]

Why do we need to use them at all? What the heck is GIGO?

Garbage In

Garbage Out

Pergametum Init

Exit Pergamentum

jb

 

[Green Maned Lion]

Oy gevaldt. Das isht meshugahss.