When debugging a system is there any generally favored approach to which bugs to start with? Is it the easiest to fix, the hardest or the "most important" (whatever that means)?
Personally, I always do the easy ones first on the grounds that in the past I have discovered that they sometimes have knock-on effects, and fixing them can get rid of others. Or conversely, trivial bugs can mask more serious ones.