---

meta: true

keywords: []

date: 191120

---

**Session 191120**

{hhr}

we were talking in length about how P Systems as a base configuration of the project would mean we depart from a "computational model" and not "an algorithm"; we discussed whether these two "things" can be thought in separation, or whether they engage in an infinite recursive containment. As a comment on that, see for example the paper Pérez-Jiménez and Romero-Campero, "P Systems, a New Computational Modelling Tool for Systems Biology" (2006; DOI).; here we find multiple *algorithms to describe the execution (simulation) of the P system* (before even implementing "something" as an algorithm within the P system model).

Perhaps useful to look at at:

- Susan Stepney, Programming Unconventional Computers: Dynamics, Development, Self-Reference

{hhr}

I also feel that today "revealed" the problem with thinking about "an algorithm" as the glue of the four layers of the piece. Say we used K-means partitioning, a random choice of "an algorithm". So if we cut off all context of the algorithmic embedding, how do we expect something common to appear from the four of us? I thus question the initial situation, and thus have intuitively come back and back again to the "contextual" descriptions, e.g. the "descriptive embedding" of for example the growing neural gas, or the P systems. That is to say, would we not create stronger cohesion by departing from the *set of objects* given by the P system formalism, for example; membranes, objects, multi-sets, adjacency, nesting, atomic time, rewriting, ... Also: If we "break down" algorithm into its ideas and context in which it was conceived, then perhaps K-means and P systems aren't really categorical mismatches (algorithm versus computational model), but they would come close to one another under the premise of "computational thinking"...?

"updating a network of homogeneous Raspberry Pi machines"