Here's an esasy one for Christmas:
The nice problem Calculating $\sum\limits_{n=1}^{\infty}\frac{(-1)^n \sin(n)}{n}$ led me to consider the sums
$$s_m = \sum_{k\ge 1} (-1)^k \frac{\sin(k)^m}{k}$$
for $m=1,2,...$
It turns out that $s_{3} = 0$.
I wonder what might be the shortest proof to make this result obvious.
We have $\sin^3(x)=\frac{1}{4}\left[3\sin(x)-\sin(3x)\right]$ and $$\frac{(1+e^i)^3}{(1+e^{3i})}=\frac{(1+e^i)^2}{1-e^{i}+e^{2i}}=\frac{2\cos 1+2}{2\cos 1-1}\in\mathbb{R}^+,$$ hence $\sum_{m\geq 1}\frac{(-1)^m \sin^3(m)}{m}=0$.
$\sum_{m\geq 1}\frac{(-1)^m \sin(m)^{2t+1}}{m}$ can be computed through the same principle. For any $k\in\mathbb{N}$, be denoting through $[k]$ the element of $(-\pi,\pi)$ fulfilling $k-[k]\in 2\pi\mathbb{Z}$, we have $\sum_{m\geq 1}\frac{(-1)^m \sin(km)}{m}=-\frac{[k]}{2}$, and $\sin(m)^{2t+1}$ can always be expanded as a Fourier sine series.