A Dev’s Accountability For Their Code and the VW Scandal
So, this popped up in the Sensei’s feed on Twitter…
And it did bring up an interesting point, that being that while it’s ridiculous to assume that a couple of developers “went rogue” and inserted code to screw up VW’s emissions, what responsibility does a developer have to question what they’re asked to do? Could the developers in question been asked specifically to fudge the operations and results of their code?
Possibly. It can’t be discounted. And if they knew, hopefully they informed somebody, and if coders are in a similar situation, one would hope they’d have an ethical framework to do something positive about it or at least refuse to produce the code.
But this also begs the question: was the situation such that the developers could tell they were manipulating data falsely? If the developers are handed a spec or data chart, for example, and it says these set of inputs mean you should output this, how is a developer supposed to know that what they’ve been handed will mean they’re deliberately skewing data? Even if a developer has the necessary science and engineering acumen in the specific field for which they’re writing code, a lot of times in those kinds of environments, it isn’t nearly so clear cut how the inputs your code is getting ties to specific physical data, especially in embedded devices or data acquisition. Add in that a lot of auto manufacturers use proprietary chips and systems in such applications and you’ve got an area when often the developers have to explicitly trust the information they’re given to produce their code.
I agree completely: if a developer absolutely knows they’re being asked to produce in essence a lie, then act ethically. However, let’s also not imply that it is or should be clear cut to figure out something is wrong.