In the past century, several conceptual and technological breakthroughs produced the digital computers and open the digital information age. At the very same time, the Watson-Crick model of the digital coding of the genetic information was developed. Despite this parallel development, biology as long focused in the understanding of existing systems shaped by natural evolution whilst computer science has built its own (hardware and software) objects from scratch. This situation is no longer true: the emergence of synthetic biology opens the doors to the systematic design and construction of biological (fluid) machines. However, even if fluid machines can be based on a kind of digital information processing, they differ from the discrete dynamical systems we are used in computer science: they have a dynamical structure. In this paper, we stress the parallel between the development of digital information processing and genetic information processing. We sketch some tools developed or appropriated in computer science that can be used to model and specify such fluid machines. We show through an example the use of MGS, a domain specific language, in the proof of concept of a "multicellular bacterium" designed at the 2007 iGEM competition.