Skip to content

History

Revisions

  • Add a page for mixing stems mix[flow] had no page at all. This is the mixer: which lanes you get and why a four-source model has no empty ones, what mute, solo and the faders do, the zoom gestures and how far they go, and the one-time engine install the first separation does. Co-authored-by: Codex <noreply@openai.com> Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01DwCZhhdrJEPTXfXEgmHunc

    @ryangavin ryangavin committed Sep 1, 2026
    bd9f270
  • Document the Xenon graticule form Co-authored-by: Codex <noreply@openai.com>

    @ryangavin ryangavin committed Sep 1, 2026
    d491102
  • Document the Xenon vault form Co-authored-by: Codex <noreply@openai.com>

    @ryangavin ryangavin committed Sep 1, 2026
    6ce6e40
  • Document the Xenon meridian form Co-authored-by: Codex <noreply@openai.com>

    @ryangavin ryangavin committed Sep 1, 2026
    bfd7807
  • Document the Xenon spindle form Co-authored-by: Codex <noreply@openai.com>

    @ryangavin ryangavin committed Sep 1, 2026
    36b3201
  • Document Xenon rosette and corolla forms Co-authored-by: Codex <noreply@openai.com>

    @ryangavin ryangavin committed Sep 1, 2026
    14a0b2d
  • Document Astrolabe Xenon flows Co-authored-by: Codex <noreply@openai.com>

    @ryangavin ryangavin committed Sep 1, 2026
    9c55096
  • Document the rounded Gyre form Co-authored-by: Codex <noreply@openai.com>

    @ryangavin ryangavin committed Sep 1, 2026
    81dbfdd
  • Document the Xenon armillary family Co-authored-by: Codex <noreply@openai.com>

    @ryangavin ryangavin committed Sep 1, 2026
    2f08a54
  • Document the counter-wound Rotor form Co-authored-by: Codex <noreply@openai.com>

    @ryangavin ryangavin committed Sep 1, 2026
    dec5688
  • Document the cuboid Truss form Co-authored-by: Codex <noreply@openai.com>

    @ryangavin ryangavin committed Sep 1, 2026
    dcf381b
  • Document the eased Weave tumble Co-authored-by: Codex <noreply@openai.com>

    @ryangavin ryangavin committed Sep 1, 2026
    b7bc031
  • Document the Iris form Co-authored-by: Codex <noreply@openai.com>

    @ryangavin ryangavin committed Sep 1, 2026
    e76760e
  • Document woven crossings and relief arcs Co-authored-by: Codex <noreply@openai.com>

    @ryangavin ryangavin committed Sep 1, 2026
    05f2855
  • Document relief and loom forms Co-authored-by: Codex <noreply@openai.com>

    @ryangavin ryangavin committed Sep 1, 2026
    212499b
  • Document woven and orbital forms Co-authored-by: Codex <noreply@openai.com>

    @ryangavin ryangavin committed Sep 1, 2026
    12046f5
  • Document connected trace visuals Co-authored-by: Codex <noreply@openai.com>

    @ryangavin ryangavin committed Sep 1, 2026
    c9aaacd
  • Say that light goes past white, and that copies get dealt The node manual regenerates itself, so `vary` and the two new modes arrived in Nodes.md on their own. Visuals.md needed the prose: a neon filament is driven above what the screen can show and `spread/bloom` is what puts that excess back as halo, which is the difference between a lit line and a pale one. `shade` gains `filament` — one colour from unlit to blown — because ramping across the palette paints hue onto geometry and a curve whose colour changes along its length reads as a thermal image. And `vary` is here, with why a copy number wired straight into brightness gives a ramp round the ring rather than sixteen separate lights. Twenty-four flows now: `depth` and `comet` join the five. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01SGPm8ZaGY63UWZRkGg2feU

    @ryangavin ryangavin committed Sep 1, 2026
    1e45ae5
  • Name the five flows made of glowing line The library table said seventeen and lists twenty-two now. The last five are one family rather than five unlike each other, which is worth saying in the guide along with why: they are the only place the drawing nodes are written down as something you can take apart, and what they are not is a folder of video files — the loop is the beat, the colour is your colourway, and there is no resolution they were mastered at.

    @ryangavin ryangavin committed Aug 31, 2026
    a1eab27
  • Add the drawing nodes to the manual, and the oscillator merge to the guide The node reference is generated, so `figure`, `array`, `glow`, `shade` and `form` arrived in it on their own. The guide is not: it calls out the nodes worth a paragraph, and five new ones with nothing said about them would send a reader to a reference page to work out what a distance is for. So: how a shape becomes a lit line, why `array`'s second outlet is the point of it, and what `form` costs — it brings its own light rather than being bloomed, and turning `chrome` up trades the glow for a reflection of a room built out of the colourway.

    @ryangavin ryangavin committed Aug 31, 2026
    814f28d
  • Say what the device face says now, and admit the page can lag The device stopped counting attached windows a while ago: Status reads `Connected to Live` whether or not an app is on it, and underneath is a row per app — set[flow], visual[flow], chart[flow] — with a dot that lights while that one is connected, and `plus N more` for anything else attached. The old wording described a connection count that no longer exists, so `No connections` was being offered as a diagnosis for a state the device cannot report. Both pages now carry a note that they are hand-written and can fall behind, and that Live wins the disagreement. It is the honest thing to say on a page describing a face that changes with the build and only changes here when somebody remembers — the same drift that put eleven `lens` modes on Visuals until the reference was generated.

    @ryangavin ryangavin committed Aug 31, 2026
    e65940a
  • Every node, every mode, every inlet — generated from the rig Visuals has a table of what you can drop on a canvas, written by hand and read before you start. This is the other half: the thirty nodes in full, with each one's modes, what every inlet is for, and the value it starts on until you say otherwise. It is generated — `npm run dev:node-manual` in the repo — from the same descriptions the app prints on the node, so the manual and the app cannot come to disagree about what a mode does. The list it joins had said eleven `lens` modes since the twelfth was added.

    @ryangavin ryangavin committed Aug 31, 2026
    3cca25e
  • Say where the apps come from Installing walked from the device straight to "open set[flow]" without ever saying where set[flow] comes from. It is a `.dmg` on the same release now, so say so, and say what a damaged-app dialog actually means: a build from somewhere other than a release. Co-authored-by: Codex <noreply@openai.com> Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_016mXxJ6RwpjXVuFbD94wF1L

    @ryangavin ryangavin committed Aug 28, 2026
    95d6228
  • Document the lineage forest Co-authored-by: Codex <noreply@openai.com>

    @ryangavin ryangavin committed Aug 28, 2026
    a4613fd
  • Document the visual Archive Co-authored-by: Codex <noreply@openai.com>

    @ryangavin ryangavin committed Aug 28, 2026
    9ea8522
  • Document visual Finals Co-authored-by: Codex <noreply@openai.com>

    @ryangavin ryangavin committed Aug 28, 2026
    dab22d6
  • Document recursive visual search Co-authored-by: Codex <noreply@openai.com>

    @ryangavin ryangavin committed Aug 28, 2026
    1bf0be3
  • Document Electron HMR development Co-authored-by: Codex <noreply@openai.com>

    @ryangavin ryangavin committed Aug 28, 2026
    a5fde89
  • Document the read-only visual Examples scheme Co-authored-by: Codex <noreply@openai.com>

    @ryangavin ryangavin committed Aug 27, 2026
    f3fc0c7
  • Document evolutionary visual training Co-authored-by: Codex <noreply@openai.com>

    @ryangavin ryangavin committed Aug 27, 2026
    ff89445