Hacker News new | past | comments | ask | show | jobs | submit login

I bet it boils down to an intersection of a number of issues.

> The solution seems trivial enough.

Maybe so, but that is the usual trap, isn't it? Us armchair experts can expound to no end from the outside.

> Write down the customer id, date, time and what was pre-approved on a piece of paper. Then when the computer comes back up, enter it into the system.

I imagine that the typical phone-drone has no authority to provide such an authorization, and with the system down they have no avenue to pursue one. No trail through the system, no audits, etc.

> Did the developers think that their system was infallible and would never go down?

Surely not, but the developers knowing their system is fallible is a long way from management acknowledging that reality, and drafting policy to account for outages. Allowing, e.g., for manual pre-auth procedures.

> If so, what was the motivation of the domain experts? [...] Or do they take delight in finding loopholes where they can shirk their contractual obligations?

Or do they get tired of dealing with bureaucratic red tape, scope changes, approval processes, etc. etc. and just ship the product the customer asked for, despite shortcomings in the spec, and call it a day so they can gtfo?




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: