rdf - The difference between blank nodes and variables in SPARQL queries -

I have studied this topic and have not found it interesting too. Although the definitions are very complex, I still do not get the answer to my question.

I can not find any examples of queries with empty nodes, which gives different results from that query that are in place of variables, examples of empty nodes.

For example, when anybody extends different results to the following questions, then any case is:

1.   SELECT? A? B WHERE {? A: Define _: Empty node _: Blank node: other project? B }    
    Select 
    ? A? B WHERE {? A: predicate? Variable? ? Variable: other projects? B }     
    Perhaps there are more complex questions that behave differently?  
    Especially I wonder if there are any examples of different results  
    thanks.  
     PS Yes, I know that empty nodes can be used only if it is in contrast to the variable in a basic paragraph but this is not the difference about which I am talking about.   
    
    
    The answer you have added is empty, in the query that the data is not being asked about empty nodes Nodes are completely correct that the variable specification, such as empty nodes in the query, says (it has been emphasized):  
    
     
     empty in graph pattern Nodes act as a variable  No, data is being sought in the context of special blank nodes.  
    Empty nodes are either labeled as "_: abc", or abbreviated "[]". An empty node that can be used in a single location in the query syntax can be indicated with []. A unique blank node will be used to create a triple pattern. The empty node label is labeled as "_: abc" for the empty node with the label "ABC". In the same question, the same empty node label can not be used in two different basic graph patterns.   
    As such, your query  
     SELECT? A? B WHERE {? A: Vicious _: Empty node _: Blank node: other project? B }   
    ? A? B WHERE {? A: predicate? Variable? Variable: other projects? B }   
    The behavior of using blank nodes instead of one variable is that you can use some more compact syntax. In this case, you can write:  
     SELECT? A? B WHERE {? A: predicate [: otherPredicate? B]. }   
    Actually, in this case, because you are only looking for a property on the thing that matches the empty node, you can use a property path:  
     SELECT? One? B. WHERE {? A: predicate /: otherPredicate? B. }