You are viewing a single comment's thread.

view the rest of the comments →

0
4

[–] FlintRockBone 0 points 4 points (+4|-0) ago 

That wouldn't happen in any professional development environment I've ever been a part of. Anything involving critical sub-systems has a code review before it's accepted into the stable source control branch. And if one rogue engineer can push code unnoticed into the stable branch then you have a poor software release process. Although maybe I've just been spoiled by being on teams that take software stability and security seriously.

1
2

[–] Gargilius 1 point 2 points (+3|-1) ago 

...out of curiosity, not questioning your creds, but how many years have you been working in this industry?

0
2

[–] FlintRockBone 0 points 2 points (+2|-0) ago 

about 12 years. The only time I was in a situation where one rogue engineer could push code into a branch we were going to release (without triggering a code review or QA analysis) was a two developer startup. I've know plenty of people who didn't have that kind of situation and they invariably suffered from software quality problems. I've always worked in desktop software so maybe there is a different attitude when you are working with physical devices, but I assumed they would have tighter policies since peoples lives depended on their software behaving in a predictable manner.

0
0

[–] TheFutureIsBright ago 

What you describe literally will not happen in development processes of critical parts of automotive software, in Germany or any other first world country. Car software that can kill thousands of people and ruin your company is not created in death marches. If anything it was discussed to death, planned out carefully and N managers signed off on it.

0
0

[–] Adynatos ago 

I can only speak about one consumer electronics megacorp, but it totally happens. What's more, there were extensive systems to show who commited some patch, some rarely used (and often circumvented because there was no time) systems to show who did a code review (usually some other engineer), but there's no trace of managerial decisions leading to that patch. Someone tells team leader, who tells project leader, who tells engineers. There's no git for managers and the company acts to shift the blame to the smallest cog possible. I think that if VW goes with it and some happless dev is blamed, they will demoralize their current employees and turn away possible new ones. They will lose more than money.

0
0

[–] FlintRockBone ago 

That's true, I always try to make make email into git for managers, but they always try to shift sensitive conversations to informal venues. One thing that I've done in the past in add a comment to the code that I don't entirely agree with saying per discussion with XYZ or something. I've mainly done that so that future maintainers have someone to hate other than me though.