In the post WWW with Metabolism and The chemical connectome of the internet was evoked the idea of using the chemical concrete machine not on biological, wet networks (although I still find this line of research very promising), but on the www. I am very happy to see that efforts concerning the use of chemical programming and bio inspired models of programming already exist!
Tell me more about this, if you read this and you are an expert! (And, of course, excuse my ignorance, I am a geometer, not a CS expert, therefore I shall mention your previous work here, as I learn about it.)
Fraglets is an extremely promising direction: here is the fraglets site and here is a fascinating article by Christian F. Tschudin. The title of the article is “Fraglets – a Metabolistic Execution Model for Communication Protocols” and the abstract reads:
In this paper we introduce a molecular biology inspired execution model for computer communications. The goal is to lay the ground for automatic network adaption and optimization processes as well as the synthesis and evolution of protocol implementations. Our execution model is based on the unification of code and data, featuring a single unit called “fraglets ” that are operands as well as operators. We have built a simulator and started to program classical communication tasks with fraglets that show metabolistic pathways patterns like the ones found in biological cells. We give an example of a fraglet implementation for a non-trivial flow–control–with–reordering protocol and briefly discuss how to search the fraglet program space with genetic algorithms.
I like very much two things here: “unification of code and data” and the “metabolistic” word in the title.
Lidia Yamamoto is another researcher who works in the (finished?) BIONETS collaboration, with Biochemically inspired emergent computation, co-authored with Thomas Meyer. Also, with Christian Tschudin, “A metabolic approach to protocol resilience” (here, from p. 191).
Of course, Banatre is the first who introduced the concept of “chemical programming” and Berry with Boudol the CHAM, or the chemical abstract machine. (Recall that the “chemical concrete machine” denomination points to the CHAM, but it is “concrete” because really it works with “concrete” molecules, involved in “concrete” chemical reactions, without using any name (or name management), and without the need for evaluation in the computation).