"When and why" to do "what it does"

This might sound reductionistic, but computers are there to: show information, do stuff to information, get information from people/hardware/internet and to store information.
And one important aspect of computers is thus the "when and why" to "do stuff". Usually a computer would be sensible and "do stuff" at appropriate and sensible times, by definition, I suppose. So when is it sensible for a computer to do things?

• When you tell it to. Through some user interaction, via a user interface, you tell a computer to take some information and "do stuff" to it
• When a certain condition has been met. In other words, when there is a cause. For example:
• when a certain time occurs, or
  
• when a certain thing happens, or
  
• when a certain, possibly derived, value is attained
• That is about it I think. Unless your computer does something for no causal reason, in which case, it is an intelligent computer, or otherwise, even non-deterministic, and thus not very smart to talk about today.
  

So what is the point?

The point is that if we can enumerate all possible "when and why" a computer will do something, we can separate the "when and why" from the "what it does", by delegating the "when and why" bit to the operating system to manage. The "when and why" for user interface could range from:

• command line commands as triggers
• mouse clicks
• voice commands
• any form of input really

This decouples the action from the trigger, allowing a more flexible and separated system. Although at the cost of speed.