Tonight was yet another in a long series of failed new architecture deploys at one of my clients. To those of us with experience in the field, this was no surprise. At two something in the morning, the CTO said "Why couldn't tonight have been boring?"
Ever the person to tell the emperor that he is naked, I replied: "Well, when you push code that has never made it through a two hour live test into a production environment you have to expect it to blow up."
Being the least experienced person in the room, the CTO replied: "That is just the pessimist in you."
Damn... Gotta show the emperor a mirror... "No, that is the part of me that has run several development projects from beginning to end, and that used to lead a QA team. That's the part of me that saying that."
I don't think he was looking at me as he actually said: "It doesn't have to be that way."
I give: "Last I heard the great and powerful wizard of oz was no longer on the throne."
While it is true that coding can be an art, the steps used to test and verify code are in fact a science. You can build test cases. You can apply logic. You can reproduce results. It wouldn't be Computer Science if you could just wish it didn't have to be that way, and click your ruby slippers together chanting "this code will run" three times. Arthur C. Clarke said "Any sufficiently advanced technology is indistinguishable from magic", and by that extension anyone who thinks they can wish code into running is not particularly advanced themselves...