91–100 of 160 posts
Background: I used to be an automotive software engineer. I speak for myself only here. Whenever the topic of automotive software comes up on HN there are comments alongs the lines of "global variables bad", but not much construtive feedback. I want to explain some of the tradeoffs that lead to the architecture used in automotive software to get a better discussion going with HN readers about that architecture. tl;dr…
Earlier quoted context omitted.
While I certainly don't buy the "pesky engineers" argument, it's actually somewhat plausible Porsche didn't know about the cheating (beyond what everyone knew, which was that the stated mileage and emissions were virtually impossible to achieve by any known mechanism). Since they dropped a VW manufactured engine (3.0 TDI) into a shared VW chassis (Touraeg) and built the Cayenne around that, there's a good chance the…
Why would a set of engineers make such a decision? This is clearly a decision which should be taken only by the SEO? What could their motivations be?
So... the engineers comply.
I once had an engineer come to me saying that he was being victimised by other engineers about a technical solution. I took my most trusted engineer and asked him to independently review the code. He came back and said the code was crap and needed to be rolled back. We rolled it back, a code walkthrough was performed on this code and we went into testing. When that function failed, I approached my most trusted engineer. My translation of what he said it "I didn't like the guy"... So, I've learned that even great engineers can be compromised, despite many claims I see on the Internet.
Regarding Porsche's knowledge, I agree with another poster in this thread who said that Porsche may have tested the unit in the conditions that didn't show the emission problems. Without further information, i'd give Porsche the benefit of the doubt.
Earlier quoted context omitted.
That's why I'd like to get it tested. I think they may have gone with the 2016 since that's what is currently on the lot, but I am very skeptical it is new to the 2016 model year when the required standards and engine did not change.
Why would you want to get it tested? Why not just wait for the EPA to test another 2015? I'm not sure what information benefit you'd get from commissioning your own test that would offset the time and money it would take to execute such a test.
Background: I used to be an automotive software engineer. I speak for myself only here. Whenever the topic of automotive software comes up on HN there are comments alongs the lines of "global variables bad", but not much construtive feedback. I want to explain some of the tradeoffs that lead to the architecture used in automotive software to get a better discussion going with HN readers about that architecture. tl;dr…
Earlier quoted context omitted.
Apparently a lot of people who make these sorts of modifications can pass emission tests. They hook up a settings swapper to the ECU, and when they need it to be clean, they put it into factory default mode. It is essentially the same sort of cheating as VW.
On a scale that isn't even remotely as large as what VW has done. The scale matters, a lot. You're talking about something affecting maybe tens of thousands of trucks, versus 11+ million vehicles including Porsche. It doesn't make sense to make the comparison.
The scale of an impact of one really bad car when it comes to certain emissions shouldn't be understated; it is important to remember that VW was simply seeking pollution they could get away with. Emissions tests are done on individual cars rather than a bulk standard by make/model for that exact reason.
VW is the second largest auto-maker in the world...few comparisons are going to hold up if that's how you think.
But what about used ones? Seems like prices will be plummeting. And if you like the cars, seems like a great time to buy.
Background: I used to be an automotive software engineer. I speak for myself only here. Whenever the topic of automotive software comes up on HN there are comments alongs the lines of "global variables bad", but not much construtive feedback. I want to explain some of the tradeoffs that lead to the architecture used in automotive software to get a better discussion going with HN readers about that architecture. tl;dr…
Quality post. Thanks for sharing. If some of these timers are implemented as separate OS tasks that can interrupt each others, what's to stop one from taking a bit too much time and throwing off the timing of another?
Earlier quoted context omitted.
If X person(s) modified their X car(s) in such a way, they'd be forced to take it/them off the road. The value of X should not matter.
I think it does matter. Think of every VW, Audi, and Porsche dealership and manufacturing plant in the US being mothballed overnight. Not even mentioning the secondary effects, that's a lot of unemployed Americans that are paying for mistakes made by people they've never met.
And? So the company structures its mischief so that any correcting action will damage innocent bystanders, therefore holding everyone in check. What happened to "America doesn't negotiate with terrorists"?
Background: I used to be an automotive software engineer. I speak for myself only here. Whenever the topic of automotive software comes up on HN there are comments alongs the lines of "global variables bad", but not much construtive feedback. I want to explain some of the tradeoffs that lead to the architecture used in automotive software to get a better discussion going with HN readers about that architecture. tl;dr…
> The microprocessor has to run for 10+ years in a wide range of temperatures and be dirt cheap, so you end up with specs like 180 MHz, 4 MB of flash and 128 KB of RAM.
If I'm paying tens of thousands of dollars for a car, how come they're using the cheapest possible components? If tinkerers can ship the Raspberry Pi for $30 per board, inc. a 900Mhz quad-core chip and 1GB RAM, you'd think GM could get components at least that modern for an insignificant cost relative to the car they are controlling.
Earlier quoted context omitted.
On a scale that isn't even remotely as large as what VW has done. The scale matters, a lot. You're talking about something affecting maybe tens of thousands of trucks, versus 11+ million vehicles including Porsche. It doesn't make sense to make the comparison.
I was comparing methods, not scale. The scale of an impact of one really bad car when it comes to certain emissions shouldn't be understated; it is important to remember that VW was simply seeking pollution they could get away with. Emissions tests are done on individual cars rather than a bulk standard by make/model for that exact reason. VW is the second largest auto-maker in the world...few comparisons are going t…
The EPA's goal is to regulate the overall quality of the atmosphere. To the extent that they regulate individual machines, it's just because of their contribution to the total.
When something is done by a few idiot hobbyists, it's just not worth trying to enforce. The effort would be better used elsewhere.
Individual cars are tested for emissions not so much to catch illegal modifications (although I'm sure they're happy to do so when they can) but to detect failures due to age or broken equipment.
"Rolling coal" is obnoxious and ought to stop, but it shouldn't be surprising nor upsetting that there isn't much enforcement effort against it.