There was a recent discussion in our department about the expectation that users of the espresso machine clean up after themselves. The discussion went on way to long and culminated with this:
"But let us think about this for a moment, in a scientific way: If n people each brew m cups of coffee a day and the machine needs to be cleaned every k cups (where we assume that it so happens that water needs to be refilled whenever the dump box needs to be emptied), then the machine needs to be cleaned nm/k times a day; or dnm/k times over a period of d days. Furthermore, as d increases, each of those n people asymptotically has to clean the machine (dnm/k)/(nm) = d/k times a day. Interestingly we observe, that each person has to clean the machine d/k times a day, whether she/he walks away without cleaning it or cleans it another time when she/he wants to brew another coffee. Of course this assumes everybody always cleans the machine either after or before usage and that somebody who finds the machine uncleaned does actually clean it if her/his intent was to brew a coffee.
You are proposing to clean the machine after usage because everybody likes to have a clean espresso. However, then we need to rely on people actually cleaning it after usage. Here I propose a more robust solution: If everybody cleans the machine before usage (if required), then everybody is forced to clean the machine if she/he wants a cup of coffee. If somebody is not willing to clean it, she/he is inherently "punished" by not getting a coffee."
The danger of letting computer scientists discuss cleanliness.