When software is built, the question that is usually asked is: how do I make it work?
But I feel there is a way more important question we need to start asking once the basic functionality is there: how does it break?
If we want reliable, safe, secure, resilient software, we need to fully understand its failure modes.
But the incentives in the industry are aligned against that. It works – ship it! Move fast and break things! First-to-market!
And so we find 24 year old dumb bugs.