I logged a bug this weekend on a project I've been working on. The bug was pretty simple: under certain circumstances, when you clicked on item n in the list, you would see a few of the children of item n-1. You could tell because they still had the icon from item n-1 next to each of the children.
And a patch was applied in under two hours.
I poked at it and discovered that I was still seeing children in the wrong place. Only now, they didn't have the icon.
So I emailed the developer to be sure that he had actually fixed it. His answer....
"No, but I hid the error"
I laughed for a few minutes, then I started to tease him mercilessly, then I stopped.
You see, this is demo software. His fix didn't fix anything, but it did prevent an error from showing up in the demo. And when the demo is scheduled for Monday morning, sometimes hiding a bug is okay. Eventually it should be fixed, but considering the circumstances, I have to agree with his call - the low risk incurred to hide the problem was worth it; the higher risk to fix the problem wasn't.
Given a problem, we can live with it, change it's behavior (hide it in this case), or fix it. Any of the three are options, and we'd do well to remember that. Depending on what's going on, the right answer might be any one of those three choices.