deploy5-desktop-rudder.png is replaced with a screenshot of Rudder running (Ready, with a
transcript). It carries no names, emails, paths, hostnames or tokens, and its metadata is
stripped: the PNG holds only IHDR, IDAT and IEND. The alt text says what it shows.
delegation-model.svg, re-rendered to delegation-model.png at 2x (only IHDR, IDAT, IEND):
- development, not main, is at the centre, behind a landing lock that takes one landing at a
time; the plugin ring and the three outputs are kept (the desktop output working).
- A coordinator and a gate-keeper replace the single overseer; the legend gains them.
- "How a change lands" has four steps: own branch; a second reader; one at a time (in the
lock: merge development, rebuild, run every check, push only if nothing fails); promote
to main (only when the owner approves, after full checks on that version).
- The outputs band adds: the dev preview follows development, the published site follows
main.
README, the same section, so the text matches the picture: the summary table row, the
section's intro, the image's alt text, the kitchen list (now five points) and the glossary
("landing lock" replaces "merge queue"). No tool or personal names.
- The clone examples in README.md and documentation/DEPLOY.md name a private server; they
now read `git clone --depth 1 <your-repo-url>`.
- skills/plugin-designer/references/backend.md quoted a planning heading that names a
platform; it now says the tiers are recorded "as an open, unproven assumption".
- The desktop app (Tauri; Rudder runs as one) works: the capability table's Desktop column
is ✅, and the outputs table, the Output 3 heading and its screenshot's alt text say so.
The WordPress column keeps 🧪, so the table's legend is unchanged.
- delegation-model.svg: the desktop output's label reads WORKING, in the green of the other
two outputs. delegation-model.png is re-rendered from it at 2x in Chromium. Rendering the
previous SVG the same way reproduced the committed PNG with 0 pixels different, and the
new render differs only inside the desktop box. The PNG carries only IHDR, IDAT and IEND.
main gained one commit, 4a0fe87d0 (README.md only), so development no longer contained it
and the promotion could not fast-forward. The edit meets development's rewritten BLUF, so
README.md conflicted. Resolved as development's README with the same edit applied:
- "- **Why:** To Bless the Nations" is the section's first bullet, then a blank line, as
in main;
- development's existing Why paragraph is relabelled "- **Great Commission:**", its text
and line breaks unchanged.
Nothing else changes: README.md is the only file that differs from development.
The shadow-styles id must be a plain lowercase identifier (check:shadow-css-parses
reads anything else as a CSS rule concatenated onto the id), so 'PanelSheets'
becomes 'panel-sheets'; nothing imports the component yet. The derived library
index gains panelSheets.js and PanelSheets.vue.
bun run capability-exemplars on the current tree. The posterverse bundle landed
without regenerating this table, so check:capability-exemplars (gates:post)
reported that the table disagrees with the manifest. Doc only; no bundle.
blessPoster.js now plans its 5 x 8 ft sheets, finished size, screen layout, glue band, trim line and sheet
tags through library panelSheets.js instead of its own copy. The print HTML, the screen layout and the plan
are unchanged: hashed for 3 presets and the default in all 11 locales (44 cases), before and after, all
identical; a 0.001 in change to the glue label moves all 44.
check:panel-sheets (gates:pre, headless) pins both geometries (60x96 stacked, 66x66 side by side, one
sheet; glue and bleed):
- the sheets tile the finished poster;
- glue slices tile it, and bleed slices repeat exactly the join;
- a bleed join prints a guide, not a solid band, and a screen preview always shows the red band;
- the Bless 5x8ft takes its bands and marks from the module.
It is two-sided: 4 planted defects are caught, and the Bless arm fails on the old private copy.
A poster too big for one sheet prints as several, each trimmed edge laid over the next sheet's join band:
60x96 is three 60 x 36 in sheets stacked with 6 in joins (the Bless 5 x 8 ft wall), 66x66 two 36 x 66 in
sheets side by side with a 6 in join (India). library/components/poster/panelSheets.js is pure:
- planPanels(sheets) takes the size table's sheets object as is (sheetW, sheetH, count, direction, joinIn),
plus the renderer's borderIn and joinFill;
- sliceFor(plan, s) gives the part of the finished poster that sheet s prints, one page per sheet;
- screenLayout, panelChromeHtml and panelCss give the sheets pulled apart and the marks.
A 'glue' join leaves the band blank and marked red; a 'bleed' join lets a map continue under it, so a
trim a few millimetres off shows map, never a white seam, and only a guide is printed there.
PanelSheets.vue previews it for Vue surfaces:
- a src (the finished poster as one image) sliced per sheet with CSS, or a #sheet slot;
- it fills its container's width, takes no pointer events, and emits one load or error per src;
- it always draws the red glue band and the gap, even for a print that bleeds.
Its label is poster.glue_here, new in all 11 catalogues.
A second country without a sheet size would silently join India's oversized
series and print India as "map 1 of 2". The comment now says so, and
check:country-sheet-one-path fails when more than one country has no sheet,
naming them, so that change is a decision rather than drift. Its self-test
plants a second one (China) and requires the arm to fail.
The wall had its own inline wrapper around generateCountrySheet, a second
copy of @map/packet/countrySheet.js, which the Regions tab uses. Two copies
of one wrapper drifted twice (a teal Regions print, then India's series). The
wall now calls makeCountrySheet with the same country list, still handing it
the data's own English record rather than the localized card.
Behaviour: identical sheets for every country with records; a country whose
records come back empty now yields no sheet (the wall falls through to its
stored preview), as the Regions tab already did.
check:country-sheet-one-path (in gates:pre, with a self-test) parses every
plugin source: a direct generateCountrySheet call fails, every
makeCountrySheet call must pass the series list, both known callers must be
found, and seriesFor gives named answers over the real data (India 1 of 1,
each 3x3 country N of 173, an absent code no series). It fails on
development c4655072a for all three defects.
seriesFor skipped any country without a sheet size, so India's packet page 1
from the Regions tab printed "map 1 of 0", while the Country Posters wall's
own inline copy printed "map 1 of 1". A null sheet is now a tier like any
other. A country absent from the list (a host's code with no poster) still
gets no series, so the renderer keeps its default rather than a made-up
position.
BUNDLE=posterverse vite build, then bun run tree and fix:sri. The manifest
gains the posterverse bundle only (plus the bundle and profile counts); the
Research Gallery's CORS-readable copy stays byte-identical to it.
<posterverse-app> is a vertical-tab rail of whole poster plugins, each mounted
as its real element from the manifest: UPG 100 List Posters (upg-100-mfe),
Country Posters (country-posters-mfe), Poster Design, and an Experimental
folder of Wall of People Groups, All People Groups, UPG Neighbor and Bible
Posters. It reuses VerticalTabsLayout, loadScriptOnce/whenTagDefined and the
Research Gallery's profile-config contract (urlSync:false, mapTitle, token
ladder, treeBase, assetsBase).
The whole rail is an allow-list (src/tabs.js), so a gated plugin such as
key-cities can never appear. Rows whose bundle is not in the tree are pruned;
Poster Design and Bible Posters carry provisional ids until their bundles land.
meta.json declares nested-mfe-spa and composes the five rows that exist, so
the Plugin Studio lists it under capability 5.
check:posterverse-shelf (in gates:pre) asserts the allow-list, the key-cities
exclusion, composite-not-leaf resolution, composes/rail agreement and
11-locale string parity; --self-test plants one defect per arm.
d3-zoom on a canvas reports its transform in the canvas's own CSS px, but the renderer applied it inside
the "meet" fit (a leftover from the SVG, whose viewBox made it world units). Every pan therefore moved
the wall by only the fit scale: a 200 px mouse drag panned 34 px at 1440, a 120 x 200 px finger drag
14 x 23 px on a phone, and a wheel zoom drifted about 90 px off the pointer. The transform now applies
after the fit, in Bless and Social mode alike.
Also: a trackpad pinch (ctrl-wheel) zooms the wall instead of falling through to a browser page zoom,
with a capped step; and a fling can no longer leave an empty table (the wall always covers the centre
of the view). check:wall-pan (gates:pre) runs the real WallCanvas methods with d3's transform
arithmetic and catches the old formulas.
check:poster-wall-host-seams gains arm H: a default wall's broken stored
webp retries the tree's own copy before the placeholder. Arm I: a generated
preview that fails shows the placeholder at once. The gate serves embedBase.js
as if from a built tree, because the second base is derived from that URL and
is empty from source, which let a mutant that skipped the retry pass.
previewSrc's observer now uses the wall's own scroller as its root (the
viewport when the wall is embedded and the page scrolls it). A pane wall
scrolls inside .wall, which clips its cards, so with the viewport as root the
200px margin never looked ahead. Arm E asserts the root.
PosterWall states the two limits of previewSrc: a timeout ends the wait, not
the render, and the page cache never revokes the URLs it holds.
A card's thumbnail was always the stored previews/<id>.webp. A gallery of
poster templates drawn from the library has no stored file, and must show a
library change the moment it lands, so the thumbnail can now come from the
host: previewSrc: async (item) => an image URL | null.
Default null keeps the stored webp and creates no observer, so every current
wall is unchanged. When set, a card renders only once it is in view, one
render at a time, cached on the page by instanceId + item id. Until it answers
the card shows its label; null, a throw or no answer within 20 s shows the
existing "preview unavailable" placeholder.
The header-action gate becomes check:poster-wall-host-seams and covers both
seams: arms D to G for previewSrc (default webp and no observer; lazy, one at
a time and the null fallback; the timeout; the cache). Six mutants of the new
code each fail it.
The home link joined Collapse's rule (.vtl-collapse, .vtl-home { … }) so the
two share one box, and the gate only recognised a rule that named
.vtl-collapse alone, so it found no reference and failed on development. It
now reads the reference from any rule whose selector list names
.vtl-collapse as a whole selector (not :hover, -tx or a descendant), and its
self-test gains a poison that moves only that rule's radius.
A template card (Poster Design) carries no people-group count, and the new
branch printed a bare ' UPG ·' before its label. The count and separator now
render only when item.count is set, in the headerAction branch only; the
default toggle is unchanged. check:poster-wall-header-action gains arm C for
it, which fails with the guard removed.
esc() prefixed a single quote to a value starting with = + - @, but OWASP's
formula triggers also include a leading TAB and CR, which a spreadsheet can
strip before reading the = behind them. Every map's CSV export shares this
escaper. The shape gate carried a formula probe but never checked it came
back neutralised; it now does, for =, TAB and CR, by slug.
The sticker on every poster card was hard-wired to "Show on map". A wall with
no map beside it (a gallery of poster templates) needs the same card with a
different action, and a fork of the card would drift from the wall.
headerAction null (the default) renders the existing toggle unchanged: the
same markup, aria-pressed, the is-last highlight and the window event, so
every current wall is untouched. { labelKey, ariaKey } renders a plain button
labelled from those keys that emits 'header-action' with the item and
dispatches no window event.
check:poster-wall-header-action (gates:pre) mounts the real PosterWall
headlessly on a recording renderer and calls the sticker's onClick in both
modes. Four mutants fail it: no seam, a new branch that also dispatches, a
default that emits, and aria-pressed on the new branch.
One more people group moved from Unengaged to Engaged upstream; the total is
still 2107. The maps compute from the live rows and already show the new
split (read on Prayer Progress's engagement legend), no shipping code
hardcodes the totals, and the live endpoint was read independently, so only
this baseline lagged.
