I'm not very good with analysis (I never studied it) but because of my "work" on other topics of mathematics I came to this problem.
$$\lim_{k \to +\infty }\log_k(k^a + k^b)=\max(a,b)$$
I'm sure that this is really "reasonable" because I tryed to graph it for really huge values of $k$ ... but "reasonable" is not enough in mathematics: I'd like to prve this. So I'm courious to know how one should go to prove it in a formal way ... if it is true (I hope).
PS: I know nothing about Limits and their rules
Without loss if generality assume $a\ge b$ then:
$$\log_{k}(k^{a}+k^{b})=a+\log_{k}(1+k^{b-a})=a+\frac{\ln(1+k^{b-a})}{\ln(k)}$$
where I changed the base of the logarithm to get the second equality. Notice that if $a>b$ then $1+k^{b-a}\to1$ and $\ln(k)\to\infty$ both as $k$ tends to $\infty$. If $a=b$ then $1+k^{b-a}=2$ but $\ln(k)\to\infty$ as $k\to\infty$. Thus, the above tends to $a=\max\{a,b\}$.