>>"For database nature loved the world so much that it gave all developers the one and only tool needed to prevent crashes(OTHERWISE), so that whoever useth properly shall not perish from their jobs"
>>
>>KG 3:16
>>
>>:)
>
>You mock a man of sincerity and strength, Kevin. You try and weaken me using various tones in your posts toward me.
>
>You are not trying to help me. You are trying to hurt me.
>
>I already said in my original post that I always code for the impossible scenario using a type of silentError() function. My question was on other people's opinions, because I've known excellent coders who held opinions on both sides. It depends on their philosophy of software. At what point do you cease assertion testing? Do you go back up and make sure that "abc" + "def" is exactly equal to "abcdef"? What if there's an error in the concatenation algorithm on a machine?
>
>It is a matter of philosophy when it's under an impossible scenario. My philosophy is and has been to code for it anyway, for the reasons you outline. So you and I are in agreement on this issue.

It was meant as a joke. Lighten up!

There's perhaps a variant of "Freudian slip" when you say, "I've known excellent coders..." Yes, have great coding chops is very important. So are business realities. I've known good coders who wouldn't always know the forest from the trees because they've never been exposed enough to both. Nothing should every "let something crash"

Now, if you want to talk about raising errors that you or other systems have agreed will be managed (in some kind of technical spec), that's a different story. Raising errors can be a valid context. It's the "letting something crash" that no serious professional should consider.

Again, it was a joke. Lighten up, dude!!! :)