A locked tile names the card it is hiding - #482
Merged
Conversation
The tile drew `???`. It draws `03 / 37` now — the card's place in the whole catalogue, so a gap is a card the learner can go and earn rather than an anonymous blank. With #396 leaving exactly one locked tile on screen, that tile is always the next one. Earned tiles gained the same line and their kind's own wash. The tint table is **generated from `CARD_TINT`** rather than transcribed: thirty-eight entries over eight hues at 8–12%, so a kind the design washes green cannot drift red here. A brewed Coffee Challenge stamps the corner; a waiting one rings it, dashed. **Two things #434 said about the design were wrong.** *The silhouettes are unreachable.* `LockedSilhouette` branches on `'silhouette'` and `'dot'`, and is called with `card.kind` — which is `botanical`, `burrs`, `roastcurve`, never either. Every real card falls through to the mono `?`. Porting three shapes would have ported two the design never renders. *The tile prints no tag.* The ticket lists "title, then the tag" on both sides and marks it as agreeing; the prototype draws only `cc-title`, and the CSS has no `cc-tag`. So the module line goes, as it went from the sheet in #385 — the card already says what it is. The test that pinned it now makes the stronger claim: no module name anywhere, since one could only be a section header. `CoffeeCardModel` carries its `kind` — the collectible's own key, which the assembly was dropping in favour of the module's icon, and which the tint is keyed by. `cardsGridItems` hands each tile its catalogue place, because the number is where the card sits in the set and not in the grid. **The artwork is #480, and it is not an art job.** All thirty-seven `CardArt*` components are static SVG taking no props, and the repo already extracts the icon family from the prototype rather than redrawing it. That ticket says so; the mark stands in here meanwhile.
…gn marks every one (#434) **The offer ring was far too narrow.** The design's `challengeOpen` (`screens.jsx:1621`) is *earned, has a challenge, has not brewed it* — so a learner sees every card that still owes them a brew. I had wired it to the single challenge the lifecycle happens to have active, which rings at most one tile and blinks off when its window lapses. It reads the bank now, and drops the `activeChallengeProvider` dependency entirely. My test had pinned the narrow behaviour, so the fix failed it: it described what I built rather than what the design asks for. **I rewrote a border the repo already had.** `DashedRoundedBorder` exists for exactly this. Mine broke the `ShapeBorder` contract four ways that no test would have caught: `copyWith` dropped the side it was handed, `scale(t)` returned `this` so any lerp jumped, no `==`/`hashCode` so the decoration never compared equal and the box repainted every rebuild, and the ring stroked *on* the radius, putting half the line outside the shape. The design writes a bare `1px dashed` and names no pattern, so the corner takes the app's own dash rhythm rather than a second invented one. **A layering inversion:** `challenges/domain` imported a `cards/presentation` file to learn `CardChallengeState`. The enum is a domain fact — none, open, tried — and now lives in the challenges domain, with presentation importing it. `total` folds into `PlacedCard` instead of travelling beside it to every call site, and `formatCardPlace` moves to `cards_grid.dart` with the rest of the set arithmetic, where a unit test reaches it without pumping a widget. Both design claims this branch rests on were verified independently: the silhouette branches really are unreachable, and the design's tile really prints no tag. The generated tint table was checked entry for entry. Recorded, not silently dropped: the `VISUAL GUIDE` top line replaces `CARD NN` for a kind no shipped collectible has.
27 tasks
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Closes #434, whose body now carries the two corrections below and the artwork ruling.
Needs your sign-off
The module tag under each tile title is gone. #434 ticked that line as already matching the design — it does not: the prototype's tile renders only
cc-title, and the CSS has nocc-tag(screens.jsx:2447,index.html:704). ADR-0009 backs removing it, and #385 removed the same line from the card sheet for the same reason, but it is a user-visible change the ticket said was fine as it was.Two things #434 got wrong about the design
The locked silhouettes are unreachable.
LockedSilhouettebranches on'silhouette'and'dot', and is called withcard.kind—botanical,burrs,roastcurve, never either. Every real locked tile falls through to the mono?. Porting three shapes would have ported two the design never draws. Verified independently in review, against both the prototype and the shipped bank.What review should look at
The offer ring marked one card where the design marks every one.
challengeOpen(screens.jsx:1621) is earned, has a challenge, unbrewed; I had wired it to the single active challenge, which rings at most one tile and blinks off when its window lapses. Fixed in the second commit — and my own test had pinned the narrow behaviour.I rewrote a border the repo already had.
DashedRoundedBorderexists; mine broke theShapeBordercontract four ways no test would catch (copyWithdropping its side,scalereturningthis, no==, and the stroke half outside the shape). Deleted.The tint table is generated from
CARD_TINT, not transcribed — 38 entries over eight hues at 8–12%, checked entry for entry in review.What defers
The artwork is #480, and it is not an art job. All thirty-seven
CardArt*components are static SVG taking no props, andtool/extract_icons.jsalready bridges prototype SVG into bundled assets. Nobody has to draw anything; the extractor needs sentinels for the illustration palette. TheVISUAL GUIDEtop line is recorded too — no shipped collectible has that kind.Checks
format, analyze, metrics clean; 2214 tests. Rebased onto the ADR renumbering pass.