Hacker News Reader

Comment by neya | Hacker News Reader

While the article does mention one more guy (Oliver Schmidt), it's a shame they just put the whole thing on this Engineer, when in reality there is no way this could have been pulled off by two people without the rest of management's knowledge/approval.

Are you seriously telling me that only this single guy was responsible for all the emissions that were found in the TDI range of diesels, all the way from A3 to A8? Sorry, I'm not buying it.

View on HN

This is a common sentiment here on HN, and it is entirely devoid of factual grounding.

This is one guy who happened to travel to the USA (stupid move: asking US authorities before travelling "hey, are you going to arrest me?" and taking their word that no, they don't plan to…). Another American VW guy has already been sentenced.

The "real culprits" are all living in Germany. And the investigation in Germany is ongoing. It may not make a splash in the US press, but there have been raids at VW offices and in managers' private homes.

Former CEO Martin Winterkorn is amongst the now thirty-seven people investigated (so it is certainly not the case that German prosecutors focus on a few lowly engineers as "fall guys", as HN commenteers love to claim).

The investigation is ongoing, and it is generally expected that there will be criminal charges. But this takes time.

Press release by the public prosecutor: http://www.staatsanwaltschaften.niedersachsen.de/startseite/...

VW and the German government are basically "cooperating" on everything [1] [2] [3]. It's extremely naive to think that VW will get a fair trail in Germany.

[1] https://qz.com/1045619/germanys-diesel-scandal-shines-a-ligh...

[2] http://www.climatechangenews.com/2017/09/05/lobbying-data-re...

[3] https://www.nytimes.com/2015/10/02/world/europe/germany-volk...

The German government doesn't control the prosecution. Especially not because it's not a federal investigation. Criminal matters are state matters, with very few exceptions.

And to counter your next try: No, the government of Lower Saxony does not control the prosecution either.

government doesn't control the prosecution but government makes the laws and regulations. that's enough. And government in Germany is super cozy [1] [2] with car makers.

[1] https://www.nytimes.com/2017/08/02/business/energy-environme...

[2] https://www.bloomberg.com/news/articles/2017-07-18/germany-w...

Last time I checked the prosecution was bound by instructions (§ 147 GVG), in this case by the government of Lower Saxony. Of course there are some limits to this and I'm not saying anything like that happened here.

Additional fun fact: The state of Lower Saxony owns 20% of Volkswagen AG (https://www.mf.niedersachsen.de/themen/beteiligungen/volkswa...)

> But this takes time

If processes against German bank managers serve as example, it will take exactly long enough to satisfy the statute of limitation.

So, in short, these guys are being charged because they can. They're a victim of circumstances. Scape goats. Political/power pawns if you will.

Understand the reasons why, but it does feel like justice.

"everybody who has anything to do with this affront to the EPA shall burn at the stake"

Oops. I meant does NOT feel like justice.

The guy lead the emission compliance unit in the US. He was most likely in charge of taking the software designed in Germany and adapting it so that it works with the emission tests performed in the US. As far as I can remember, the software was pretty closely monitoring the driving style and comparing it to the test suite that was performed in the EU in order to activate the cheats. They probably had to perform the same kind of thing in the US, as I assume that the tests differ.

And QA engineers or security reviewers never detected this in the source code? I would love to see how many people participated in those source code repositories.

Except, they don't work with actual source code. An engineering team responsible for adaptation for certain market, most likely works in calibration space, that's it, just data (flags, thresholds and tables, thousands of them). To access this data they rely on ASAP2 description databases which are simply very large text files [1] containing ROM addresses, parameter names, etc. I don't know the whole VW workflow, but can imagine how one could define a set of switching parameters for cheating, and then delete their definitions, leaving these parameters invisible.

[1] https://vector.com/vi_datadescription_ecu1_en.html

Are you saying that once the configuration is fonr there is not a QA environment?

No, I'm only saying this calibration QA has no readable source code to work with. And there are ways to obscure some parts of the ECU behaviour.

Nope, the articles says there are 7 more higher ranking executives due to be judged. And from the looks of it their sentences might be significantly higher as well.

I don’t see any claim or implication that this is intended to represent the entire punishment for the incident.

The article implies other executives to be sentenced. Liang happens to be the lowest ranking executive. Also from the looks of Schmidt’s sentencing it seems the orher executives probably had longer sentences.

Reply on news.ycombinator.com