It’s been a bit since I’ve written much about it directly, since things keep moving around faster than it felt useful to write about them, but this evening, it feels appropriate to write a bit to plan out what is becoming known as “the Emsemioverse,” which is the shape my digital data and how i work with it is taking.
This log is written for myself and other agents working in that universe to help us understand what my goals are and how I picture getting there.
One thing I’d like to explain early on is the relationship between a few key documents, because their relations reflect my background as a writer and how that shows up in this space.
A lot of times, I might say “Emsemioverse” but what I’m talking about would genrealize out to any Semioverse. And, before continuing, readers should know what an “interactive semioverse” and “semiotic universe” are, as mathematical objects.
Style Guide, Specifications, Documentation, Runbooks, Code
In the Emsemioverse, the determining sources of truth are in the Library of Specifications. Almost all of those are in the earliest steps of their formation, but the idea is to have good specificatiosn for things like, what a math paper needs to have in order to be “well-formed”.
I’m basing these specifications loosely on RFCs, but hoping to develop each out with their own suite of examples, scripts, tests, whatever is appropriate for the given specification.
But! Critically, in order for any of those sources of truth to be meaningful, there has to be some specification for writing and words themselves. That’s where the Style Guide comes in.
Written for authoring text in general, it provides the guidance that says how to write the Specification Specification, which in turn provides the ground for every other specification.
So, the source of form is the Style Guide, which lets us express our sources of truth.
There are a few specifications which are most critical to a Semioverse, but first I want to babble about structure.
Repositories have Modules have Cells
In a Semioverse, “Semiotic Repository” is the biggest available unit. The Emsemioverse is one big repository.
It contains a bunch of Semiotic Modules. That’s what the library is full of. You can tell a directory is a semiotic module because it’ll have its own REAME.md, and probably an AGENTS.md.
And then, any given file might be viewed as a “Semiotic Cell”.
Though it’s important to understand, modules are cells. And repositories are modules, and so are cells.
Agential Semioverse
This is a concept that’s not really been discussed yet, but I’m realizing as I try and apply the concept of interactive semioverse to a real living computer system, I’m reaching for an “Agential Semioverse,” which, beyond being a better buzzword for the contemporary zeitgeist, may show what additional structure needs to be in the Interactive Semioverse to let it have the degree of autonomy I want from this space.
But, for now, we’ll just treat “agential” as an informal extension of the interactive semioverse.
Core Web of Specification Modules
Markdown
Repository
Module
Cell
Agent
Useful Modules for Doing Stuff
Pandoc
xs MkDocs