Many engineering fields (including CS in some contexts) have a huge dependency on experimental setups that are "temperamental" at best much of the time and/or dependent on someone else's experimental work and other hardware that doesn't always perform according to spec, etc.
As someone with degrees in non-CS engineering fields, I mostly find the idea that programming on a stable computer has a high frustration/grief factor to be pretty funny.
