feat: the rings split derived from invented, not who typed it - #84
Open
khaneight wants to merge 1 commit into
Open
feat: the rings split derived from invented, not who typed it#84khaneight wants to merge 1 commit into
khaneight wants to merge 1 commit into
Conversation
The outer ring held both compiled and extrapolated articles, and the page told them apart by filling one dot and outlining the other. That distinction was doing no work. Both are prose a language model wrote — splitting on authorship says nothing about either. What actually differs is whether the claims trace to a document. A compiled article sets down what a source says; that is the same act as reading a corpus into a trait, and it belongs beside the traits. An extrapolated article asserts things no source contains. So `Distilled` holds both the traits and the compiled articles, and `Written` holds only work that no source contains. Which also puts the mark where it belongs: the outer ring *is* the machine-original work, so its colour already says so and the fill-versus-outline device is gone. `persona` was the wrong name for the middle ring once it holds more than the persona — it is `distilled` now, and `work` became `written`, since everything in the archive is the clone's work. Two consequences fell out. A compiled article's `sources:` edge is *authorship* rather than a citation: the document is where its claims come from. Only an extrapolated article's `sources:` is a `grounds` citation, because following that as ancestry would credit the corpus with prose it does not contain. And `links` no longer joins fixed layers — an article in either ring may link to one in either — so its `EDGE_KINDS` entry declares none rather than asserting a pair that is not true. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Stacked on #83. You were right that the distinction was silly.
The diagnosis
The outer ring held compiled and extrapolated articles, and I was telling them
apart by filling one dot and outlining the other. That distinction did no work —
both are prose a language model wrote. Splitting on who typed it says
nothing about either.
What actually differs is whether the claims trace to a document. A compiled
article sets down what a source says; that is the same act as reading a corpus
into a trait. An extrapolated one asserts things no source contains.
So compilation moved into the middle ring
And you were right about the name too:
personawas wrong for a ring holdingmore than the persona. It's
distilled;workbecamewritten, sinceeverything in the archive is the clone's work.
This also puts the mark where it belongs. The outer ring is the
machine-original work, so its colour already says so — the fill-versus-outline
hack from #81 is deleted. The safeguard got stronger by being structural instead
of decorative.
Two consequences that fell out
A compiled article's
sources:edge is authorship, not citation — thedocument is where its claims come from. Only an extrapolated article's
sources:stays agroundscitation, because following that as ancestrywould credit the corpus with prose it does not contain.
linksno longer joins fixed layers. An article in either ring may link toone in either, so its
EDGE_KINDSentry declaresfrom_layer: Nonerather thanasserting a pair that is not true — and the test skips what is not declared
instead of being loosened.
On the demo
One purple node. Which is honestly how much original work that clone has done.
543 tests, clippy and fmt clean. CLAUDE.md compressed under budget again — this
time by moving the
exportbullet's detail todocs/publishing.md, which nowcovers all of it.
🤖 Generated with Claude Code