The exercise is:
Fix $b>1$. If $m, n, p, q$ are integers, $n>0$, $q>0$ and $r=m/n=p/q$, prove that:
$$(b^m)^{\frac{1}{n}} = (b^p)^{\frac{1}{q}}$$
Hence it makes sense to define $b^r = (b^m)^{\frac{1}{n}}$
So far, this is what I got:
Since $n,q>0$, $m$ and $p$ must both be positive or both be negative, or else the equality $m/n=p/q$ won't hold. Now, $m/n$ is not necessarily an integer, but $mq$ surely is. The same holds for $np$. It is trivial to see that $(b^m)^{\frac{1}{n}} = (b^p)^{\frac{1}{q}} \iff b^{mq} = b^{np}$.
Lemma: For every real $x>0$ and every integer $w>0$, there is one and only one positive real $y$ such that $y^w = x$.
Proof: See Rudin's Principles of Mathematical Analysis, page 10.
Hence, $mq=np$. So, indeed $(b^m)^{\frac{1}{n}} = (b^p)^{\frac{1}{q}}$.
Is this proof satisfactory? Am I missing something?
I think you have the right idea. One minor comment is that when you write “Hence, $mq=np$”, you imply that the statement $mq=np$ follows from some immediately previous statement, such as the the lemma. But it doesn't; it follows from the hypothesis that $m/n = p/q$.
But more importantly (IMO), this line
needs more explanation. When students write that something is “trivial”, I usually take it as a bluff. To me the point of the exercise is to use the lemma you cited to prove that $y^n = b^m \implies y^q = b^p$. Your highlighted statement follows.
Addendum: Based on the comments below, I think you're still missing the point of the exercise and how the lemma is to be used. You're right that the lemma guarantees $(b^m)^{1/n} = (b^p)^{1/q}$, but if I understand the exercise correctly, you are to provide the details.
The lemma tells us that every positive number has a $w$-th root, and how to recognize the $w$-th root of a number from its properties. In other words, $w$-th roots exist and are unique.
What is $(b^m)^{1/n}$? It's the unique positive real number $y$ such that $y^n = b^m$. Likewise, $(b^p)^{1/q}$ is the unique positive real number $z$ such that $z^q = b^p$. To prove that $(b^m)^{1/n} = (b^p)^{1/q}$, it is sufficient to show that $y^q = b^p$ Why? And how?