On phones (up to 767px) the "What can this map do?" panel opened as a full-screen page: a few
lines of text at the top of an empty white screen, reading as a new page rather than a panel.
It is now a bottom sheet:
- it sizes to its content, up to 70% of the small viewport (svh, so a collapsing URL bar does not
resize it), and scrolls inside, with its scroll kept to itself;
- the close button is 44px and stays in view in the sheet's header;
- the dimmed map above it stays visible, and a tap there closes the panel, as Escape does.
Desktop is unchanged. Rebuilt doxa-maps-spa (the only bundle that compiles MapInfo.vue), then
tree + SRI.
Every map whose ⓘ is its own control (MapInfo trigger 'control') now gets the same pointer:
it shows the first time that control is on screen in a page load, and goes for the rest of the
load once the reader opens the popup or leaves the tab. A full reload shows it again. This is
page-load memory, never storage, so the popup offers no checkbox in control mode: the "Keep
tooltip turned on" box and its string (all 11 locales) are removed, along with the pointerKeep
prop that selected this model per tab. Floating-mode maps keep their stored model and their
"don't point again" box, unchanged.
Leaving the tab is read from the pane (hidden or inert), on the control's intersection change
and the existing 1.5 s measurement tick.
Rebuilt doxa-simple-map (it ships the mapInfo catalogue) and doxa-maps-spa, then tree + SRI.
Prayer Progress draws its control column only once the style and the people-group data are in,
which on a slow phone can take longer than any fixed wait, and then the gallery's floating ⓘ
appeared and was swapped for the control. Now the map announces at mount that a control is on its
way ('dxmaps:mapinfo-pending'), and MapInfo waits for the map instead of a clock.
- A known failure (the map reports an error) or an unmount before the map is ready withdraws the
announcement, and the floating ⓘ shows at once.
- While a control is pending, the 30 s wait becomes a 90 s backstop (time on screen), so an
unreported failure still gets a floating ⓘ.
- A map that never announces anything keeps the 30 s wait, and pages with no MapInfo are unchanged.
Rebuilt doxa-simple-map and doxa-maps-spa, then tree + SRI.
The dashboard header gains an info button ("What can this map do for you?") to
the left of Share. Both are rounded squares at the vertical tabs' radius
(--vtl-radius), matching the app's own roundedness; map controls stay round.
Built as a reusable, parameterized header-button form so other dashboards adopt
it by prop, not by copy:
- MapControlButton: new shape="rounded-square" (border-radius: var(--vtl-radius,8px)).
- ShareButton / MapInfoButton: a shape prop forwarded to MapControlButton;
ShareButton also squares its rotating ring so a round ring never wraps a square.
Defaults stay 'round' — every existing consumer is unchanged in behaviour.
- research-dashboard DashboardTopBar mounts the ⓘ and Share as one header-button
group, sized to the bar with the 44px tap target kept for both. DOM order
[ⓘ, Share] puts ⓘ at the logical start: left in English, right of Share in
Arabic (the flex row reverses under rtl).
- doxa-maps-spa dashboard-plus-grid tab -> trigger:'control', so the tab's
MapInfo is driven by the header ⓘ and the gallery's floating ⓘ is suppressed:
exactly one ⓘ, fail-open. No new strings (reuses mapInfo.open, 11 locales).
The control column goes back to the library's round, spaced buttons at every width; the square
grouped card read too compact, on phones most of all. The ⓘ stays the top control.
In control mode, MapInfo no longer shows its floating button before the map's control has
announced itself. The map draws its column only once it is ready (style and people-group data),
so the floating ⓘ appeared for a moment and was then replaced by the control. The wait is bounded
at 30 s of time on screen, counted only while the pane is visible and not inert: measured column
arrival times run from 7 to 19 s on a local serve, so a shorter wait expired inside a normal load
and flashed late. If no control arrives, the floating button shows, so there is still always a way
to open the popup. Once a control has announced itself, measurement alone decides, as before.
MapInfo's plain <script> block (the page-load pointer memory) now sits above <script setup>, so
check:setup-order no longer reads it as used before its definition; its explanation moves into
<script setup> so the library index keeps the component's own description. Compiled output is
unchanged by the move.
Rebuilt doxa-simple-map and doxa-maps-spa (the only bundle that compiles MapInfo.vue), then
tree + SRI.
- pointerKeep maps (Prayer Progress only) point ONCE PER PAGE LOAD: the pointer shows the first time
the map is on screen in a load. Opening the popup hides it for the rest of that load, and the next
full load shows it again. The memory is a module-scope Set (pointerDoneThisLoad), because the gallery
remounts a heavy tab's MapInfo on every switch back. It is marked when the pointer is on screen, not
on mount. Nothing about it is persisted any more (no seen write in show() or hide()).
- The keep box (still stored, per map) keeps the pointer after the popup closes and on every
appearance. A page-load pointer ignores the every-map off switch, which is not offered beside it.
- The label is "Keep tooltip turned on" (en), with de retuned to match; the other nine already say
"keep ... on".
- Rebuilt doxa-simple-map and doxa-maps-spa, then tree and fix:sri.
In the detail view a thumb on the content resized the sheet whenever the
content could not scroll that way, so reading the list moved the sheet. Only
the handle (the drag strip and the title row) starts a resize now; below it
the content scrolls, in every mode.
The sheet's list gets overscroll-behavior: contain (the list only, not the
pane), so a fling that reaches its end stops there instead of scrolling the
page behind the map.
The open state was a fixed 30% of the map: a three-row legend sat in a mostly
empty sheet, and a long tree scrolled inside one. At rest in the open state the
sheet now takes its content's height, capped at 45% of the map. A drag still
sets an explicit height; the collapsed and fully open states are unchanged. The
detail view's mini image follows the open sheet's real top edge.
The start state was a race. Every map's own mount code means collapsed on a
phone, but uiStore.init() measures the breakpoint in a ResizeObserver whose
first callback runs after mount, so Simple Research read the 'desktop' default
and started open on every load at 390, while Prayer Progress started
collapsed. The sheet now applies the collapsed start state itself, once, the
first time it sees the phone breakpoint, unless a people group is selected.
The open sheet covered the licence credit and the Mapbox wordmark (measured at
390 on Simple Research: the top element at both was the sheet's content). The
bottom corners now stay at the map's bottom edge and the sheet sits above a
strip measured from the real boxes of both engines' bottom controls.
The approximate-location caption sits in that strip, under the sheet. It shows
once whether the sheet is open or collapsed (it was 0 when collapsed), and the
pin-hint pill, which lifts itself clear of the sheet's top, never lands on it.
- Prayer Progress's column becomes the engine's square grouped card (controls="square"). Its first
control is MapInfoButton, which opens the gallery's MapInfo for the tab. The gallery's tabInfo
'prayer-map' entry now uses trigger 'control', so the floating button hides while that control is
on screen, and shows again (fail-open) whenever it is not.
- MapInfo, control mode: the first-visit pointer now points at the control. Its place is measured
against the pane (which scrolls) and goes on the control's physical inner side, so an Arabic page,
whose column mirrors to the left, gets the pointer on the right with the arrow reversed. It shows at
every width: the under-1200px rule was measured for the floating pointer crossing each tab's top
strip, and does not apply here.
- MapInfo, pointerKeep (prayer-map only): the popup offers "Keep tooltip activated" instead of the
every-map "don't point again" box. It is off by default, and opening the popup retires the pointer.
Checked, the pointer keeps showing on every load. It is stored per map.
- mapInfo.keepPointer in all 11 locales.
- Rebuilt doxa-simple-map and doxa-maps-spa, then tree and fix:sri.
A v-show-hidden button is still a child, so it took the square toolbar's
:last-child corners (MapToolbar controls="square") from the last button a
reader can see, on every page where no host answers (a standalone leaf, a
partner page). Found by exp reading the bridge.
The control now asks from the toolbar's own element (MapToolbar provides it
as mapToolbar.el), or from its placeholder's parent, and renders its button
only once a host has answered (v-if). The rendered button then announces
itself, so the host measures and focuses the real element.
The "what can this map do for you?" button can now live IN a map's control
column instead of floating over it. This is the bridge only: nothing changes
until a map adds the control and its gallery tab sets trigger: 'control'.
- map-controls/MapInfoButton.vue (new): a MapControlButton with the info
icon and the localized mapInfo.open label. Clicking it sends a bubbling,
composed 'dxmaps:mapinfo-open' event out of the plugin's shadow root to
the gallery pane. On mount it asks 'dxmaps:mapinfo-control' and shows
ONLY if a MapInfo on the same pane answers (ack), so a standalone leaf or
a partner page never shows a button that opens nothing. A MapInfo that
mounts later broadcasts 'dxmaps:mapinfo-host' and the button asks again.
- info/MapInfo.vue: new prop trigger, 'floating' (default, unchanged) |
'control'. In control mode it listens on its own pane: it acts only on
events whose composedPath contains the pane, and stops them there. Its own
floating button and pointer hide ONLY while the announced control is
MEASURED usable: connected, a non-zero on-screen rect, and in no
inert/aria-hidden subtree. It is re-measured by IntersectionObserver and
ResizeObserver plus a 1.5 s fallback timer, never inferred from an unmount
event. Escape and close return focus to the control when it is on screen,
else to the floating button. aria-expanded is kept on the control.
Proved on a UPG-100 pilot served locally (gallery page, 1440, MapLibre and
Mapbox, en and ar), exactly one info button on screen at every step:
wall 1 floating; Map 1 control (it wins its hit test and sits inside the
toolbar column); opened from the control, Escape closes it with focus back
on the control; toolbar hidden 1 floating, shown again 1 control; tab away
and back 1; back to the wall 1 floating. 0 page errors. That pilot's wiring
is not in this commit.
Reported from a phone: on the Regions tab, "2 posters per row" for the country
posters did not work -- a 390px phone showed ONE per row.
THE CAUSE is a deliberate clamp in PosterWall. The per-row setting is 2, but the
grid is Math.min(perRow, fits) with fits = floor((viewport - 36) / 210), and at
390 that is floor(354 / 210) = 1. So the setting said two and the phone silently
showed one.
★ WHY THE CLAMP EXISTED, AND WHY IT IS SAFE TO RELAX NOW. It was added when a
two-up card (~162-170px) CLIPPED its action labels -- 'Prayer List' and 'Poster
Map' both read 'Pra...', and Romanian's 'Pachet de cartonase' was measured at 54
of 81 controls clipped at ro/390. That was then fixed separately: the labels WRAP
(.pfs > .pbtn, "a label too long for its box wraps"). With wrapping in place the
clamp no longer protects anything; it only costs the phone a column.
★ IT IS OPT-IN, NOT A GLOBAL CHANGE. PosterWall is shared by twelve files,
several of them frozen experimental maps. A new minCardPx prop defaults to 0,
which means the existing 210px clamp, so every wall that does not set it lays out
byte-for-byte as before. Only regions-list sets 150, which gives two per row at
360, 390 and 414, still one at 320 (which genuinely cannot hold two), and no
change on desktop (already capped at the per-row setting of 2).
MEASURED on the regions-list leaf page, this branch served locally, against
FINAL on :4611 -- by hit test and by eye:
FINAL @390 en 1 per row, 367px cards
AFTER @390 en 2 per row, 169px cards, 0/123 labels clipped
AFTER @390 ro 2 per row, 0/123 labels clipped <- the old clipping corner
AFTER @390 ar 2 per row, RTL, 0/84 labels clipped
AFTER @1440 en 2 per row, unchanged
0 horizontal overflow and 0 console errors in every cell. The print buttons now
stack three high on a phone card instead of sharing a line.
KNOWN, COSMETIC, NOT CHANGED: at two-up the card's "53 UPG · Show on map" pill
wraps "Show on / map" onto a second line. Readable, not clipped.
LegendMobile already shows "The dot marks where this people is found, not a town
or a boundary." once, floating above the sheet, and placed so that it takes no room
from the options inside. The SemanticTreeLegend it renders inside the sheet printed
the same sentence again, so with the sheet open a phone reader saw it twice.
Measured on Prayer Progress at 390 on the shared browser, with the sheet opened
through its own toggle (pp-caption.mjs):
- before, on development as served by DX Life: 2 visible copies on / and on the
Deep Research tab, the floating note plus the tree's note inside the sheet;
- after, Prayer Progress rebuilt locally with this change: 1 visible copy, the
floating note, in en and ar;
- desktop, 1440: still 1 copy, the desktop legend's own tree, unchanged.
The sheet's tree now gets :pin-note="false". The detail view keeps its own note,
which carries the shared-point wording.
Library only. Every bundle that renders LegendMobile needs a rebuild to ship it.
PinHint's first-visit pill sits at z-index 2147482000, so with MapInfo's scrim at 1000 the pill floated
bright and clickable over the dimmed backdrop while the popup was open (gospel; measured on the live
build: the pill's centre hit the pill, not the scrim, on Cities). The scrim goes to 2147483000; the
dialog is its child, inside the scrim's stacking context, so it stays on top of the scrim. The root fix
is the top layer (<dialog>.showModal()), carded.
The pill's clearance above the licence credit is measured in syncAttribClearance,
and that runs again only when the credit RESIZES. On a phone the credit is also
MOVED, lifted above the legend bar, without resizing. The clearance then stays at a
stale value, and the pill covers the credit. MEASURED at 390, on the shared
browser: on DX Life's home page the pill covered 36px of the 37px credit with the
clearance at 43px; on the leaf in ar, 36px.
clampIntoView now takes the credit's measured top as a floor, next to the fold and
the phone legend sheet. It already runs from the settle timers (1.2, 3 and 6 s)
and from the sheet and host observers, so it reads where the credit is, whatever
the clearance says.
Measured on a served tree, Prayer Progress rebuilt with this and the 120px cap,
at 390: the leaf and the gallery tab, in en and ar.
- before (development's tip): overlaps of 7, 36, 7 and 7px;
- after: 0px in all 4, with the pill's bottom 13px above the credit's top.
Library only. Every ProfileLoader consumer needs a rebuild to ship it.
The clearance is the strip from the credit's top to the box's bottom edge, capped
so a mis-measure cannot throw the pill up the map. The cap was 72px. On a 390px
phone MapLibre's credit wraps to two lines, and on the leaf it sits above the
bottom bar, 90px up. With the selector now finding it, the clearance hit the cap
and the pill still overlapped the credit by 6px, in en and ar (was 35px). The cap
becomes 120px: the measured need fits, and a runaway is still bounded.
Library only. Every ProfileLoader consumer needs a rebuild to ship it.
The "What can this map do?" pointer is anchored beside the ⓘ and extends ~213px
INWARD, across whatever the tab puts in its top strip. Worst overlap per width,
measured on Deep Research on the shared browser, on 99c22caf7:
390 Regions "Region Dashboards" 1416px^2 UPG-100 1428px^2
768 Cities legend 6560 Gospel legend 6560 + input 6240 Regions 4200
900 Cities legend 2432 Gospel legend 2496 + input 2176 Regions 3696
1024 Cities CLEAN Gospel CLEAN Regions 2100
1200 CLEAN CLEAN CLEAN
1440 CLEAN CLEAN CLEAN
★ IT IS NOT A PHONE PROBLEM, which is how it was first reported and how I first
fixed it. It persists to 1024 and only clears near 1200, so the max-width:600px
version I wrote first would have left 768, 900 and 1024 broken on four tabs.
Two things found that: gospel counted PANELS (.gm-legend/.gm-sheet) as well as
controls -- a control-only probe misses a legend entirely -- and upg100 measured
the same shape on their own tab (768 over "Prayer Resources" and "Map", 1024 over
"Map", clear at 1200+), in en and ar.
★ ARABIC IS MEASURED, not argued. My first attempt used ?lang=ar, which does not
switch the locale -- the page stayed dir=ltr and I reported that run as invalid
rather than as an ar pass. Via DX Life's /ar/ route (dir=rtl confirmed) the
pointer is hidden and clean at 390, 768 and 1024, and at 1440 it mirrors to the
start side (x=104) and is clean there too.
★ TWO OTHER SHAPES WERE MEASURED FIRST AND BOTH WERE WORSE. Stacking it below the
button traded one collision for five: Gospel's "Resources found" chip (2010px^2)
plus the zoom +/- of Prayer Progress, Big Picture and Simple Research. Hiding
only its text shrank it to 30px but still left 675px^2 on "Region Dashboards" and
900px^2 on Gospel's search input.
⚠ NOT A PLACEMENT CHANGE, and an earlier draft of this got it wrong. It also
pinned the button's inline side at narrow widths; studio's pos.narrow values have
since landed and the buttons now sit at x=8, 12, 294 and 346 across the seven
tabs at 390. A pinned inset-inline-* here wins by specificity and would have
silently flattened every one of them. The button is placement's business; this
rule is only about the hint. 0 added lines declare --map-info-side; 0 touch
.mi-btn.
Its own block rather than the existing 600px one, which also full-screens the
dialog and has nothing to do with this.
VERIFIED AFTER THE CHANGE, en and ar: pointer hidden and 0 overlaps at 390, 768
and 1024 on Cities, Gospel, Regions and UPG-100; present and clean at 1200 and
1440.
NOT FIXED HERE: at 1440 Big Picture's pointer lies over the Religion filter by
1050px^2. Stacking it below the button there does not clear it -- it relocates it
to "Adopted" 561 + "1%" 425. The only single rule that clears that too is hiding
the pointer at every width, which deletes the affordance where it does fit. So
BP-1440 wants a per-tab placement value or a known-minor ruling; it is
pointer-events:none, so the filter still opens.
syncAttribClearance() measured the strip it must clear from Mapbox's credit
(.mapboxgl-ctrl-attrib) and our own (.doxa-attribution), never from MapLibre's
(.maplibregl-ctrl-attrib). MapLibre is the default engine. With no match, the
clearance stayed at the 20px fallback. On a phone the MapLibre credit wraps to two
lines, and the "Tap a coloured dot" pill covered it: a licence-required credit
hidden on every map that mounts the hint.
Measured at 390x844 on DX Life's home page (Prayer Progress), on the shared
browser: MapLibre en and ar, pill over the credit by 35px with the clearance at
20px; Mapbox, clearance 72px and 0px overlap.
The selector now also takes .maplibregl-ctrl-bottom-right .maplibregl-ctrl-attrib
and the MapLibre form of our own credit line. The existing measure (the strip from
the credit's top to the box's bottom) and its ResizeObserver then apply unchanged.
Proved on the live DOM, inside the map's shadow root: on MapLibre the old selector
finds nothing and the new one finds the credit; on Mapbox both find it.
Library only. Every ProfileLoader consumer needs a rebuild to ship it.
The button and its pointer were placed with a physical right:12px, so on an
Arabic page, where a tab bar starts at the right, they covered the first
(active) tab. They now use inset-inline-end / inset-inline-start (LTR is
unchanged), and under dir=rtl the pointer's arrow glyph and its nudge are
flipped so it still points and moves toward the button. --map-info-top and
--map-info-side remain the per-tab overrides.
The shadow card behind a packet's first page sits at z-index:-1; without a
stacking context of its own it was painted under the dialog's white
background and never seen. isolation:isolate on the stack keeps it just
behind the page.
MapInfo (library/components/info) answers "what can this map do for
you?" with real output the map makes, pre-generated so it opens at once.
The examples are an ordered list of kinds rather than a fixed set,
because maps make different things: an image (a poster), pages (a packet
that shows its first page and expands to every page), a deck (cards that
expand from a stack to a grid) and a table (a real excerpt of the
spreadsheet a map exports, with translatable headers). An example whose
image is missing drops out quietly: no broken-image icon, no empty frame.
The pointer at the button goes for good once that map's popup has been
opened and closed; a checkbox in the popup turns it off on every map.
Both are kept in localStorage behind a guard, so private browsing cannot
break the page. On a phone the popup takes the whole screen; Escape
closes it and focus returns to the button. The pointer's motion stops
under prefers-reduced-motion.
All of its text is one new catalogue area, 'mapInfo', in all eleven
locales: the component's own words, and each map's lead, takeaway and
note as mapInfo.tab.<tabId>.<field>. One area because a bundle ships only
the areas its source names, and the gallery that mounts a popup per map
names none of the maps' own areas. This commit carries the UPG-100
map's two sentences; the other maps' follow.
RegionSectionCard passed THREE arguments to useShadowStyles(css, label): the CSS,
then 'regions-list', then 'region-section-card'. The signature takes two, so
'regions-list' was the key and its own name was silently ignored.
useShadowStyles keys its <style> tag by label and, on a collision, OVERWRITES
the existing tag's text instead of adding one (useShadowStyles.js:60-61). This
card is a CHILD of regions-list, so the two share one shadow root: the cards
render after the fetch, mount second, find the list's tag under the shared key
and replace every .rl-* rule with the card's CSS.
Measured in the running page: .rl-root became position:static and .rl-scroll
overflow-y:visible, the shadow root held 3 style tags where a working tree holds
4, and 'rl-root{' appeared nowhere in it. The list therefore had no scroller of
its own, a wheel moved the pane column instead, and every region below the fold
was unreachable -- at 1440 Middle East, North America and Oceania; at 390
everything except Africa.
It was latent until the label became real: before the cyrillicHeading fix, a
dangling comma concatenated 'regions-list' INTO regions-list's own CSS, so the
list keyed by a content hash and the two never met. Fixing that rule turned a
silent collision into a visible break, which is why the live site is unaffected
and every sha after it is not.
Verified in the artifact, not the source: ,"regions-list") appears once in the
rebuilt bundle where it appeared twice, ,"region-section-card") now appears once,
and the .rl-* rules are still present.
BoundedSet.vue, CenteredSetTarget.vue and VisionFunnel.vue came back through the
handoff merges, byte-identical to what c5c01dead deleted on purpose on 09-28
("Delete 7 unreachable library components"). That commit also added
check:no-dead-library precisely so they could not return. Verified before
removing: 0 files import charts/BoundedSet.vue, charts/CenteredSetTarget.vue or
charts/VisionFunnel.vue, and packetCharts.js references none of them, so nothing
is left dangling. check:no-dead-library passes with them gone.
A first count said BoundedSet had 6 importers. It has none: those files import
layoutBoundedSet from charts/bounded-set.js, a different module, and the grep
matched the substring.
packetCharts.js: the doc block describing scatterCells sat above topCells' own
block, two functions away from what it documents. Moved to sit directly above
scatterCells. No code change.
The handoff merges brought in three references that name the Driver in shipped
text: two comments in CenteredSetTarget.vue and one _why in the people-groups-grid
instances. Reworded to state the engineering reason without the speaker -- an
open decision now reads as pending a product decision, which is what the rule
asks for. The instances file still parses as JSON.
Clearing the clip is not the same as being clickable. The attribution and the
pin hint are painted over the canvas and take pointer events, so a popup that
ends fully inside the frame can still have its buttons underneath them.
Measured at 390 on a 366x732 map: the pan cleared the clip exactly (cut 0) and
a real click on Full Profile still did nothing, because at y~701 the top element
at that point was div.maplibregl-ctrl-attrib-inner; on the sibling pin it was
span.pin-hint-pill. Both pins panned, both looked fixed, neither was.
Neither the popup nor the control is at fault. Attribution is a licence term of
the basemap, so raising the popup over it is not available to us. The usable
frame is simply shorter than the container, and the fit test has to say so.
panPlanFor now takes a bottomInset and fits against containerH - bottomInset.
On the measured case the plan changes from -77 (pan up, popup below the pin,
buttons under the attribution) to +119 (pan down, popup above the pin, clear of
the band) -- the planner flips direction because with the band reserved, fitting
above became the smaller move. That is the existing "pick the smaller move" rule
doing its job on a corrected frame, not a new rule.
The inset is MEASURED from the live controls rather than assumed. A constant
would be wrong on the next control we add, at a width where the attribution
wraps to two lines, or when the gallery's clearance transform lifts the band.
Only controls that intercept pointer events are measured: the corner container
is pointer-events:none and spans nearly the full map width at 390, so measuring
the container would over-reserve the whole strip.
Unchanged where it was already right: a pin whose anchor fits still returns 0
(rega measured pan 0 on those in both arms), a popup taller than the usable
frame still refuses rather than jumping the camera, and bottomInset 0 is
byte-identical to the previous behaviour.
7 probe cases, with the post-pan fit asserted from the geometry rather than by
re-calling the planner -- a self-consistent oracle cannot see an error that
shifts both sides.
With no resolver configured, the kit chip bar joined a multi-value filter with ', ', so "Africa,
Asia" read as ALL to most people and carried English commas in every locale. The selection bus
matches several values on one facet with includes() (useSelectionBus isFilterActive), so adding a
value WIDENS the result: they are alternatives. The chip now uses formatList(value, { type:
'disjunction' }): "Africa or Asia", "Africa أو Asia", "Africa或Asia". gospel traced the matcher
and made the call as the only consumer's owner (gospel-internet's GiCombos). The English on this
chip changes from a comma to "or".
check:legend-mirrors lifts this exact expression and drives it with no resolver. It now lifts
formatList from library/i18n the same way (the shipped grammar, not a copy) and expects
"Islam or Hinduism". Poisoned with the old join it fails; restored it passes. formatList's doc now
says to choose the type by what the list MEANS, and never 'unit' for names.
check:approx-pin-note server-renders ApproxPinNote in Node, and it crashed: with no locale argument
and no runtime locale, formatList fell through to detectLocale(), which reads `document`. The page
lookup is now a guarded last resort. ApproxPinNote also passes its sentence's own locale (the
host's vue-i18n when installed, else '' so the runtime/page chain applies), so a host with its own
i18n can never get an Arabic sentence with a list in another language.
The gate now expects the new English ("Koraga and Maila", the list grammar this change exists for)
and adds the Arabic case: two names joined with و, never an English comma inside Arabic prose.
Poisoned by putting join(', ') back, both of those cases fail; restored, it passes.
ApproxPinNote built pins.sharedPoint's {names} with join(', '), so the Arabic sentence carried English commas between the group names: the case gospel measured. It now uses formatList, which joins by the locale's own grammar (ar 'A وB وC', zh 'A、B和C', en 'A, B, and C').
DataTable's sheet heading fell back to a hardcoded English 'Detail', one word and so outside check:hardcoded-english's reach. It now reads aria.rowDetail ('Row detail'), which this component already uses for the dialog's own label: translated in all 11 locales, and it brings no new catalogue area into any bundle (a dashboard.* key would have added the whole dashboard area to every bundle that shows a table).
rega's poster-path timings: A4 300 dpi ~42 s, 24x36 in at 150 dpi 90-135 s,
24x36 in at 300 dpi ~4 min. Through all of it PosterDialog showed an
internal phase name ("hidden-map"), a bar measuring only the current phase
(so it restarted and sat still), and English. A failed render, including a
canvas too large for the device, went only to the console: the status block
vanished and the reader was told nothing.
- The bar follows the composable's overall, monotonic percent; beside it,
the time elapsed (ticking) and, while stitching, "tile n of total".
- An honest expectation for the chosen size, from its megapixels: "up to a
minute" from 5 MP (A4 300 dpi, 8.7 MP), "a few minutes" from 15 MP (24x36
in 150 dpi, 19.4 MP, and up). A4 at 150 dpi (2.2 MP, 11 s) shows none.
- A failure is shown in the dialog: a specific message when the device ran
out of memory (an oversized canvas or encode), a general one otherwise,
nothing when the reader cancelled (AbortError).
- The Render button's labels and the "(optional)" placeholder go through
T() too. 9 poster.* keys in 11 locales, using the dialog's own terms.
RegionCountryMap's note on the projection override said who ruled it out, in capitals the reference
baseline's pattern never matched; it now says what was decided and why the globe stays. The
mobile-gating header and a people-groups-grid census entry carried internal card ids that named their
source; the ids are shortened to what they describe.
RegionCountryMap.vue and check-mobile-gating.mjs change in comments only: script, template and styles
are identical before and after, compared without comments and without renaming. The census card field
is informational; nothing matches on it.
Defence in depth for a bug already fixed at its source. <doxa-country-map> was
passing center: null when its host set none; Vue applies a prop default only for
undefined, so that null replaced this component's [20, 10] default and reached
the Map constructor, where the engine's LngLat.convert threw before a style or a
tile was requested. Twelve measured cases, every one a dead map reported as "a
map-token or connection problem".
The element now sends undefined and no longer produces that null. These guards
are for the NEXT caller: a component in a library cannot assume every host
resolves an absent value the way one of them does.
The asymmetry is why it happened at all. `bounds` two lines below already had
`|| undefined` at the constructor, so a null bounds was harmless; `center` had
no guard, so a null centre was fatal. Both now behave the same way.
flyTo gets the same treatment, because it reaches the engine's LngLat conversion
by another path and would throw identically on a re-frame. It falls back to the
map's current centre, so a caller that supplied none gets the zoom re-framed
without the camera sliding sideways.
