Question: Why does the exception indexing function prevent me from getting an existing table entry?

Here's a very simple example, working as intended:

 

 

Now, if I try to use the exception indexing function:

 

 

We can notice that b[1] isn't using exception anymore (as if the fourth input overwrote the second one) and that b[1][2] isn't linked to the value "two". However, if I define the tables first, I get the expected result:

 

 

Why does the exception indexing function prevent me from getting an existing table entry in the second case?

Please Wait...