Suppose we create a card named Joe's Coffee+hours.  It's name contains a "+", therefore it has a compound name. 

When we have a card with this name, we know that we have two other cards: one named Joe's Coffee, and another named hours.  If you went to create Joe's Coffee+hours but there was no card named hours, then it would be created automatically.

If we were to create a new card named Joe's Coffee+menu, it would share the same left part (Joe's Coffee) with our original card, Joe's Coffee+hours. Similarly, if we created The Bumpy Grind+hours, it would share its right part (hours).

This means that we can use naming patterns alone to start creating networks of information.  With the above examples, Joe's Coffee  is "plussed" to three other cards, and hours  is "plussed" to two.

By following naming patterns like these, we can begin to organize card information in highly flexible ways.

Note that if we were to rename either part (Joe's Coffee or hours), the name of the compound card would be changed to use the new part name. If, however, we were to rename the compound card, the part cards would be unaffected. For example, renaming Joe's Coffee+hours to Joe's Coffee+menu would mean the compound card is no longer related to the hours card and is now connected to the menus card; the hours card would otherwise be unaffacted.