Another way? What does (s)he want? Hmm ...
Let and let .
The substitution shows that .
The substituion shows that and so .
Thus by symmetry.
Now put so that , for and , and .
Wait a moment, this is merely a disguised version of the method. It's no better than NonCommAlg's rejected solution.
(Thinks: Perhaps I should have used an infinite series, or some kind of limit, or introduced a parameter, or ... (Mind boggles at this point))
An answer that is not really an answer? This is trickier than it looks.