Today I got involved in a discussion at work about a feature the client wanted to implement for a particular scheduled process.
Basically, the client wanted to be notified via an e-mail when the scheduled process was finished. Perfectly reasonable.
The tech manager, however, had an additional concern: what if the data acted upon by the scheduled process were not present, in which case the scheduled process might never complete? This was a serious problem, he said, and he wondered whether this hypothetical issue ought to be mentioned to the client.
Now, to be fair, yes, the right thing to do was to ask the client (which is what the team did). But what I found funny about this discussion was this: if the data relied upon by the scheduled process were not there, it would cost the company literally millions of dollars. And the tech manager’s main concern was that if the scheduled process did not complete, then it would be automatically killed by another process, which would raise a ticket for some other team in the organization to deal with.
To me, this is like planning a party, deciding it would be a good idea to have the party catered by a Chinese restaurant and then worrying, “Oh no, but what if aliens invade Earth and decide to enslave humanity, and they don’t like Chinese food?”
Whether our hostile alien overlord guests were not fans of our food selection would be among the least of our worries in this scenario. We’d be enslaved by aliens.