@justai/cuts 0.85.0 → 0.87.0

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@justai/cuts",
-  "version": "0.85.0",
+  "version": "0.87.0",
   "description": "A persona's named parts — the page shell that holds them, and the cuts themselves.",
   "license": "UNLICENSED",
   "repository": {

package/src/Toolbar.astro

@@ -349,36 +349,47 @@ import AuthMount from "./AuthMount.astro";
      So the difference stays, and it is a difference in ink rather than in box: mark ~19px, glyphs ~14.3px
      inside the same 19px. If the row is ever to read level, the way there is to grow the GLYPHS, not to
      shrink the identity. */
-  .tb-nav.tb-mark :global(img) { width: 19px; height: 19px; }
-
-  /* ── WHATEVER IS FIRST IN THE ROW READS AT 20; EVERYTHING AFTER IT AT 19 ────────────────────
+  /* ── EVERY ICON READS AT 20 ─────────────────────────────────────────────────────────────────
+   *
+   * queries: toolbar icon size · 20 · why is my icon smaller than the others · 26.67 · the mark
+   *          looks small · icon ink
+   *
+   * The founder's rule, and it is about the DRAWING rather than the box: an icon's picture hits 20
+   * on its longer side, whichever side its design naturally uses. A bolt is tall and narrow and hits
+   * 20 on its height; a folder hits it on its width. The container is whatever that takes.
    *
-   * queries: toolbar icon sizes · why is the first icon bigger · 20 and 19 · first position
+   * One number, two boxes, because a viewBox is not a drawing:
    *
-   * The founder's rule, with app.zone's own mark as the reference every icon is measured against.
-   * It is POSITIONAL, not about which icon happens to be there — the row is saying that the first
-   * thing in it is not one of the sections, and that is true whatever fills the slot:
+   *   a MARK fills its own viewBox        → a 20px box puts 20px on screen
+   *   a GLYPH sits inside 18 of 24 (0.75) → 20 / 0.75 = 26.67px
    *
-   *   a persona          `BackToZone` — the way out to app.zone
-   *   a child persona    the parent's face — the way up
-   *   app.zone itself    its OWN mark, because there is no way out of the zone from the zone
+   * A persona whose mark does NOT fill its viewBox needs its own number, and that is where the
+   * confusion came from: those marks looked small in the same box as everything else because their
+   * drawing occupies less of it. The box is derived — 20 divided by how much of the viewBox the
+   * drawing uses — so a redraw changes it, and the comment beside each one says so.
+   *
+   * (This replaced a 19-with-a-20-in-front rule that lasted an afternoon. The distinction was real —
+   * the first slot is not one of the sections — but it is said by POSITION and the -7px pull above,
+   * not by a pixel of size nobody can see and everybody can drift.) */
+  .tb-nav.tb-mark :global(img) { width: 20px; height: 20px; }
+
+  /* ── THE FIRST POSITION STARTS WHERE THE ROW STARTS ─────────────────────────────────────────
    *
-   * Written structurally for that reason. `BackToZone` sizes itself; when it is absent the nav group
-   * becomes the first child of `.tb-left`, and its first item is what a person sees first. Expressed
-   * as a selector rather than a list of ids, a fourth arrangement is right without anyone editing
-   * this file — which is what "positional" has to mean to be worth the words.
+   * queries: the first icon is further in on a child persona · navParent inset · -7px ·
+   *          the row starts in a different place
    *
-   * Measured across nine live surfaces first: `home` was 20 on eleven personas and 19 on qr, which
-   * had overridden it; the three children's parent icon was 19; auth's was 19 because I set it there
-   * an hour earlier having read the drift backwards; app.zone's own mark was 19 while being the
-   * first thing in its row. `BackToZone`'s default was right the whole time.
+   * `BackToZone` pulls itself 7px left so the row begins flush with the bar's edge rather than a tap
+   * target's worth inside it. Anything standing in the first position has the same job and needs the
+   * same pull — measured on a live child, the parent's face sat 26px from the bar's edge where a
+   * persona's way out sits at 17. Nine pixels, found by eye before any probe looked: a person moving
+   * between two surfaces sees a difference no single page can show.
    *
-   * The first slot always holds a MARK — a drawing that fills its own viewBox — so a 20px box is
-   * 20px of ink. A glyph there would need 26.67 to read the same, and if one ever lands in that slot
-   * this rule is where the exception belongs. */
-  .tb-left > .tb-nav-group:first-child > .tb-nav:first-child :global(svg),
-  .tb-left > .tb-nav-group:first-child > .tb-nav:first-child :global(img) { width: 20px; height: 20px; }
-  .tb-nav :global(svg) { width: 19px; height: 19px; }
+   * POSITION IS WHAT MAKES THE FIRST SLOT DIFFERENT — not size. Every icon reads at 20; what says
+   * "this one is the way out rather than a section" is where it sits, flush to the edge, before the
+   * group. Structural for that reason: BackToZone sizes and pulls itself, and when it is absent the
+   * nav group becomes the first child of `.tb-left` and its first item inherits the job. */
+  .tb-left > .tb-nav-group:first-child > .tb-nav:first-child { margin-inline-start: -7px; }
+  .tb-nav :global(svg) { width: 20px; height: 20px; }
 
   /* ── EQUAL INK, TWO BOXES ────────────────────────────────────────────────────────────────────
    *
@@ -401,7 +412,7 @@ import AuthMount from "./AuthMount.astro";
    * app.zone settled it, qr arrived at the identical number, and qr's own comment says why it is
    * here: *"otherwise every persona has to remember to carry it, which is precisely the thing the
    * shared shell exists to prevent."* */
-  .tb-nav:not(.tb-mark) :global(svg) { width: 25.33px; height: 25.33px; }
+  .tb-nav:not(.tb-mark) :global(svg) { width: 26.67px; height: 26.67px; }
 
   /* THE BUBBLE HANGS. Its silhouette's centre measures 1.05px below the line the other three sit on
      (39.05 against 38.0) because the tail is part of the shape and pulls the box down with it — the