1: \begin{abstract}

2: The aim of this paper is to give a short overview on error bounds and to provide the first bricks of a unified theory. Inspired by the works of \cite{AzeCor02,BolDanLeyMaz, BolDanLew1, eb&kl,AzeCor16}, we show indeed the centrality  of the \L ojasiewicz gradient inequality. For this, we review some necessary and sufficient conditions for global/local error bounds, both in the convex and nonconvex case. We also recall some results on quantitative error bounds which play a major role in convergence rate analysis and complexity theory of many optimization methods.

3: \end{abstract}