Fix the problem you can see then deal with what's left I was working for a company that made lighting equipment for touring bands. One band had decided they wa

Fix the problem you can see then deal with what's left I was working for a company that made lighting equipment for touring bands. One band had decided they wanted the biggest lighting desk in the world (this was in the early 80's). We cobbled together two of the company's latest digital desks as master-slave, and it nearly worked. Two problems remained as we got really close to the band's tour starting: a weird flickering in some of the indicator LEDs in the slave, and the output level (controlling the lights on stage) would sometimes wander around unpredictably. We had a quick look and decided that the output level thing was a major issue which had to be dealt with first. The LED flickering problem was a) minor, and b) caused by a simple clocking problem which we could deal with later. We spent 48 hours or more with no sleep, trying to fix the output level thing. Got it just about useable, but not perfect. The tour was due to start the next day so we decided to give up on the output level problem and fix the LED flickering, so that there was only one issue, albeit a bad one. Spent an hour fixing the clocking issue so that the LED's were OK, and lo, the output problem was sorted. If only we'd dealt with the simple problem first, then we'd have managed to get some sleep! And before you say that was a one off and bad luck or bad triaging or something, I've used the same heuristic more than once and it's always paid