I'm currently trying to get an approximation to work:
Denote the set of all entire functions as $H(\mathbb C)$. Now I want to show the following:
Let $f, g \in H(\mathbb C)$, $\delta > 0$ and $A \subset \mathbb C$ compact so that $\mathbb C \setminus A$ is connected. Then there exist a $\varphi \in H(\mathbb C)$ and a $n \in \mathbb N$ with
$$ \Vert f - \varphi\Vert_A < \delta \qquad \text{and} \qquad \Vert g - \varphi^{(n)}\Vert_A < \delta,$$
where $\Vert f \Vert_A = \max_{z \in A} \vert f(z) \vert$ and $\varphi^{(n)}$ denotes the $n$-th derivative of $\varphi$.
What I can use is Runge's theorem. That gives me a polynomial $p \in H(\mathbb C)$, such that $\Vert f - p\Vert_A < \delta$ and $q \in H(\mathbb C)$ such that $\Vert g - q\Vert_A < \delta$. But I don't see how I can construct the $\varphi \in H(\mathbb C)$ from that. Sometimes one can get such a function by summation of polynomials, but I don't see how that could work in this context.
I also got the following lemma to work with: I know that $\varphi^{(-n)} \to 0$ locally uniformly for $n \to \infty$. So there exists a $n \in \mathbb N$ with $\Vert \varphi^{(-n)}\Vert_A < \delta \quad$ (or any other constant I might use. $\varphi^{(-n)}$ denotes the $n$-th antiderivative is this statement). Note that this lemma solves the case where $f \equiv 0$, but I fail to generalize it. I dont see how to facilitate this lemma in the general case.
I would appreciate some hints or good ideas on the topic. Thanks in advance :)