flp on Nostr: I spent all morning defining a format similar to semantic triples that works with ...

I spent all morning defining a format similar to semantic triples that works with what I know. This is where I'm at:

```
[
20260301T151947049Z,
c4afa2e8-5110-442c-92f5-9b7ca8ff65f8,
4c396468-2347-4bb6-a11e-dfb261d130c5,
a131174a-6130-4633-a031-82cd9cc61e0f,
7a166a8b-c3a2-4a00-8a40-87e07ea4b285,
Pj1d71yjrjzzex9WDKWRtKxbhVhIavGjCISG7cGc2js
]
```
Or:
```
[
UTC timestamp,
UUIDv4 graph,
UUIDv4 subject,
UUIDv4 predicate,
UUIDv4 object,
x (public key) from a Ed25519 exported JWK
]
```

The problem with the Tim Berners-Lee / W3C version is that it is based on the WWW (naturally). I'm focused on crisis scenarios with much of my past work, and am pivoting to the log format and journal application as a vehicle for the ideas vs. simply just writing about the tech and rapid biosphere changes and socio-economic-environmental stress of current civilization (polycrisis/metacrisis). Regardless, the crisis perspective is important to me as it is an area to focus on that leverages my skills in the context of where I see future value.

I guess these are hexads, then? I like that. If I just add graph, it is much like RDF Quads. One thing I see in collaborative work, is the evolving work is based on timestamps and identity. Perhaps x. I decided *not* to use Nostr npubs, as I want it to work directly with Web APIs and Deno... the JWK format calls the pub x for the exported Ed25519 public key. I guess the hexad x would be x vs npub? OK, so sombody has an x of Pj1d71yjrjzzex9WDKWRtKxbhVhIavGjCISG7cGc2js and enters a triple. When assembling the most recent model, any changes would be reflected until the latest. Changing it back can be done by any other x with a different date. If the model is polluted because an x doesn't know what they are doing, replay all of the hexads for the most recent model.

I'm using UUIDv4, as these are all instances of an ontology. BUT, people pick their own ontology. For that matter, I could cross hexads like this:

```
[
20260301T151947049Z,
c4afa2e8-5110-442c-92f5-9b7ca8ff65f8,
4c396468-2347-4bb6-a11e-dfb261d130c5,
https://www.w3.org/1999/02/22-rdf-syntax-ns#type,
https://w3id.org/dfd#Process,
Pj1d71yjrjzzex9WDKWRtKxbhVhIavGjCISG7cGc2js
]
```

Any kind of rendering of the model wouldn't really care if the ontology was web based or UUIDs that were folksonomies. That is, if the node is a process for a data flow it would render the same whether it was from Christophe Debruyne's formal definition or x said at a certain timestamp that 4c396468-2347-4bb6-a11e-dfb261d130c5 was a process in the ontology for that particular graph (c4afa2e8-5110-442c-92f5-9b7ca8ff65f8).

That's where I'm at today, anyway.
Ivo (nprofile…vh7s)

npub1lkwmzj49a80rsh56kwme6wxysn5w3z9c58jafmvwnv6mnlh3efesfzct3n