Tell me again, #developers, why you are trusting the presence or absence of a custom #HTTP header distantly related to a #vendor we abandoned years ago in an #authentication related decision. Do you need the #SRE side of the house to dig into your code after deploying the application to a new #Kubernetes cluster with a more sensible, vendor-neutral #ingress scheme, and point out specific lines of code underlying a #linchpin endpoint that seems to be the only one experiencing this symptom?