The print productivity and medium dimensions of wide-format roll-to-roll printing systems are limited by position errors induced by step-wise medium transportation. The aim of this article is to develop a design framework for multivariable repetitive control (RC), that enables improved productivity and accuracy through continuous media flow printing. The developed technical framework for RC explicitly addresses the tradeoff between performance and modeling requirements through decentralized designs. In particular, systematic design approaches are developed that explicitly address unmodeled interaction as uncertainty, i.e., through robustness. The result is a range of solutions, including 1) independent single-input single-output (SISO) designs, and 2) sequential SISO design. Experimental results confirm that the developed RC framework in conjunction with continuous media flow printing outperforms existing approaches.