Referring to "one drop" in the above sentence, I remember Mother Teresa who was once asked by an Indian Journalist if her works in Calcutta had added some values to the welfare of humanity considering the vast nation-wide miserable conditions of unprivileged people across whole India and the Mother had replied as though her works was like adding a single drop of water in the vast ocean without which the ocean could survive well and yet it had to live without that single drop of water if she had not added that.
"We ourselves feel that what we are doing is just a drop in the ocean. But the ocean would be less because of that missing drop." -- collected (courtesy: @J-M-L)
Transition line shows the next state of the machine and the signature(s) on the transition line indicates the cause/action that is done before the state changes.
You could say that the actions are carried when entering the state, but from a programmatic perspective, they are most often best coded when you detect the event and before setting the new state.
if you code them within the case for that state, often you'll need to add some variables for checking if you are just entering the state and need to perform the action or just waiting in that state until the next event. (some can be of no consequence, like repeating a digitalWrite() to set a pin to a given value, but if you want to beep upon the transition, then you'll need to remember you have beeped, or add a BEEP stage before transitioning to the other state....
also you can arrive to a given state from different previous states, so you would possibly need to deal with what previous state you were to perform the necessary actions putting you in that state.
That's why I represent them on the transition arrows and execute that as part of the state change if (eventXXX) {} clause