Is Prod configured differently? - Prod environment architecture has some features/behaviors our test environments don’t, so guess where a bug was found? Didn’t Know That Existed - APIs we were decomisioning were used by tools devs had no idea about Customer Wanted it that way - When an api caused a cascading data failure, business wanted to know why it was so easy to cause this problem; this response was rather satisfying to hear during a miserable debug.
That’s what keeps coming to mind when I try to reply to this topic:
Developer lost interest.
Implicit/tacit requirement misinterpreted.
Lack of rest or overworking.
Interrupted during coding by testers.
Works on 10 things simultaneously.
Not enough training received.
Assumptions made across teams.
Known potential problem introduced.
They won’t do that.
Lack of accountability/ownership.
No one looks for trouble.
This one by @sebastian_solidwork nails it for me as it goes beyond the actual flaw introduced in code (e.g. an Obi-Wan error), effectively taking away at least part of the blame from the developer(s) who introduced it.
In this volatile, uncertain, complex, ambiguous world that has interdependencies interwoven all over the place, it’s probably the biggest bug should one assume there exists a single root cause.