docs: what the corpus still reports, and why it stays

Three probed findings written down rather than fixed: an outside ::marker has
no metric influence in CSS (a bulleted line is 0.12mm short of LibreOffice's),
an unbreakable run longer than the line moves whole instead of breaking at the
margin, and a typeset formula keeps its own baseline rather than its frame's —
honouring the frame measured worse. Plus the live-vs-cached index numbers.

CHANGELOG.md

@@ -177,7 +177,12 @@ merely unimplemented belongs in the list above, not here.
   drift apart along the line (identical word widths, 0.45 mm apart by the
   150th mm). A matching negative `word-spacing` was tried and reverted — it
   fixes the break but shortens every last line; see
-  `tests/render-parity/README.md`. Noted 2026-08-09.
+  `tests/render-parity/README.md`. A third case is a run with no break
+  opportunity in it that is longer than the space left (a row of leader dots, a
+  long URL): LibreOffice fills the line and breaks the run at the margin, where
+  Chromium moves the whole run to the next line. `overflow-wrap: anywhere` would
+  match it but also breaks runs LibreOffice keeps whole and shrinks a table
+  column to its narrowest glyph. Noted 2026-08-09, revised 2026-08-11.
 - A text box anchored inside a paragraph loses the vertical offset it was
   anchored by: it is a block node here, so the importer lifts it out and it
   simply follows that paragraph (measured 4.7 mm on the thesis' figure page).
@@ -196,6 +201,29 @@ merely unimplemented belongs in the list above, not here.
   editor). LibreOffice recomputes auto-grow heights on open, so a re-saved box
   keeps its content but sheds excess empty height. Shapes (rect/ellipse) export
   with fixed geometry instead. Noted 2026-07-03.
+- A list marker does not size the line it labels. Word and LibreOffice put the
+  numbering's own font on that line, so a marker whose font reaches deeper than
+  the text grows it (measured: a Courier New bullet on 11 pt Carlito text adds
+  0.031 em ≈ 0.12 mm, which is ~1 mm down a page of bullets). Root cause: the
+  marker is an out-of-flow `::before` (`editor.css`) because CSS gives an
+  outside `::marker` no position; and an outside `::marker` has no metric
+  influence either — probed, changing its `font-family` moves nothing, only its
+  `font-size` does, and that grows the line *above* the baseline where a word
+  processor grows it below. Noted 2026-08-11.
+- **Deliberate, not a defect:** an inline formula is typeset here, not placed as
+  a box. LibreOffice stores each one as an OLE frame and positions that frame
+  (`style:vertical-pos="middle"`/`"from-top"`); we render MathML, whose own
+  baseline is the typographically right one, so the frame's vertical placement is
+  ignored on import. Honouring it was tried and measured worse — every inline
+  formula in the Math Guide sank ~1 mm and the corpus went from 100 to 125
+  reported differences. It also means a render-parity run cannot compare a
+  formula-heavy page: the two sides typeset the glyphs independently.
+  Noted 2026-08-11.
+- **Deliberate, not a defect:** a table of contents shows live page numbers.
+  LibreOffice and Word print the numbers cached in the file until the reader
+  updates the index, so a document whose cache is stale disagrees with us (the
+  deputy-standards fixture caches page 7 for a heading that is on page 8 in both
+  renderings). Noted 2026-08-11.
 
 ---
 

tests/render-parity/README.md

@@ -89,8 +89,8 @@ form), the summary (a student summary: bullets, wrapped
 pictures, one table), the thesis (a 48-page thesis with charts, metafile figures and
 a German TOC) and the Math Guide (LibreOffice's own 63-page Math Guide: a title page,
 a three-page index, per-section masters, ~35 figure frames and 400 formulas) — still
-report. The thesis matches LibreOffice's 49 pages and so does the Math Guide at 63 —
-most of what is left is the engine rules below, not layout:
+report. Both run one page off LibreOffice (thesis 48 against 49, Math Guide 64 against 63),
+in each case downstream of a line the rules below cost or gained:
 
 - **A page-anchored frame's coordinates are its page's, but a frame in a *header* keeps
   the ones its zone gives it.** Both are drawn out of flow; the body one (`anchorPage`,
@@ -110,11 +110,25 @@ most of what is left is the engine rules below, not layout:
   it holds is a float, so its height is the text's alone.
 - **A page's last line is where the footer band ends.** The body margin a file's header or
   footer reconstructs is the page margin plus the zone's declared height; the zone's own gap
-  to the body sits *inside* that height (probed — `styleResolver.ts`). Getting that wrong
-  costs a line on every page, which is worth more than any single block's spacing.
+  to the body sits *inside* that height (probed — `styleResolver.ts`). The height is the
+  zone's *own* font: it is one paragraph here, so its mark struts every line of the band
+  (`blockFontSize.ts` is in the zone schema, `hfReachPx` measures the same value). Getting
+  that wrong costs a line on every page, which is worth more than any single block's
+  spacing.
 - **Chromium floors every painted border to one pixel**, so a table's rows come out
   ~0.25mm too tall each unless the excess is taken off the cell padding
   (`docs/architecture/tables.md`).
+- **A bulleted line is 0.12mm shorter here than in LibreOffice.** The numbering font joins
+  the line it labels there; our marker is out of flow and CSS gives an outside `::marker`
+  no metric influence either (probed: its `font-family` changes nothing, its `font-size`
+  grows the line above the baseline where LibreOffice grows it below). The summary
+  fixture's bullet pages drift ~1mm by their last line — that is all its `position`
+  reports are.
+- **A formula page cannot be compared.** LibreOffice draws each formula from its own OLE
+  frame; we typeset MathML, so `pdftotext` and `getClientRects` measure two different
+  typesettings and the harness splits a row whose glyph boxes disagree. Honouring the
+  frame's `style:vertical-pos` on a formula was tried and **reverted** — every inline
+  formula sank ~1mm and the Math Guide went from 100 to 125 issues (CHANGELOG).
 - **Page geometry is document-wide.** Headers and footers are per section (each is
   editable in place), but margins, orientation and format are not: they come from the
   master page governing the most body blocks (`import/odt.ts`), so a title page with its