Was coding up a new internal tool today and spent the majority of the time figuring out all the different ways that things could go wrong.
I ended up think about whether all the error checking was a safety net or a fishing net. As you do.
If it was a fishing net, I wanted the holes to be as large as possible so that I only proceeded if the inputs supplied were as constrained as possible.
On the other hand, if a safety net, I want the holes to be as small as possible so that nothing sneaks through that shouldn't.
I'm still undecided on whether I code like a fisherman or an acrobat.
[Image credit: Mark Philpott]