We all know that -128 can be represented as [10000000] as a 8-bit two's-complement number. However, it is not in accordance with the convention which bits but the most significant one are inverted and added to 1.
I am confused by the interpretations from books and the internet. Someone said -128 was a special case, so it was prescribed as [10000000]. I am puzzled until I get Randal E.Bryant's introduction in his book.
He introduced to treat as a number written in binary notation, we obtain two's-complement representation of signed number w.
When w equals to -128, its two's-complement representation is [10000000] out of question.