Skip to content

Deck detail says what the deck can be studied with - #335

Merged
jvsena42 merged 4 commits into
mainfrom
feat/deck-detail-study-features
Sep 23, 2026
Merged

jvsena42 merged 4 commits into
mainfrom
feat/deck-detail-study-features

Conversation

@jvsena42

Copy link
Copy Markdown
Owner

The four study opt-ins — Listen, Speak, Type the answer, Both directions — were settable at publish
and in the deck editor and shown nowhere else. The reader deciding whether to keep a deck was the
one person who could not find out what it offered.

Deck detail now carries a STUDY MODES row between the tags and the stats bar, one chip per opt-in
the author turned on, reusing the deck editor's own icons and labels so the row names the switches
its author actually flipped. Read-only — these are the author's decisions, not the reader's.

Two things worth knowing

Listen and Speak are folded through speechReady before they reach the screen, as the study
session and the editor already do. A deck with the opt-in on and no declared language pair offers
neither button, so a chip there would advertise a feature the session withholds. Type and Both
directions are ungated, as everywhere else.

All four chips take the primary accent, not the editor's alternating pair. Speak and Both
directions are purple in the editor — but purple on a soft purple pill is exactly what TagChip
is, and the first build put them directly under the tag row where they read as more tags. Orange
says what the deck does, purple says what it is about; the caption is the second half of that
separation.

Verified

Driven on Medium_Phone (staging): one chip, then the language pair added in the editor and the
Listen chip appearing, in light and dark; and on Pixel_Tablet at both the medium and expanded
width classes. ciCheck, detektAll, :shared:jvmTest and checkStringPlurals all pass. Full
run in journeys/RESULTS.md, with journey 26 extended to cover the row and the gate.

iOS is unbuilt — written on a Linux box, so DeckFeatureBadges.swift and its two wiring lines
need a Mac pass before they count as verified.

🤖 Generated with Claude Code

jvsena42 and others added 4 commits September 23, 2026 08:59
The opt-ins were settable at publish and in the editor and readable
nowhere else, so the reader deciding whether to keep a deck was the one
person who could not find out what it offered.

Listen and Speak are folded through speechReady on the way out, as the
study session and the editor already do: with no declared language pair
the session offers neither button, so passing them on would let a screen
advertise a feature it withholds. Type and Both directions are ungated.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
A STUDY MODES row between the tags and the stats bar, one chip per
opt-in the author turned on. Icons and labels are the editor's own, so
the row names the switches its author actually flipped.

Two things the first build got wrong on a device. The chips take the
primary accent rather than the editor's alternating pair: purple on a
soft purple pill is what TagChip is, so four of them under the tag row
read as four more tags. And the caption is there for the same reason —
without it a second row of pills is just more tags.

Read-only: these are the author's decisions, not the reader's.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Mirrors the Android row: caption, one badge per opt-in, primary accent
throughout so a badge never reads as a TagChipView. It scrolls sideways
like the tag row above it, which is the pattern the screen already uses
for a row of pills.

Unbuilt and undriven — written on a Linux box, so it needs a Mac pass
before it counts as verified.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Records the 2026-09-23 device run — phone light and dark, tablet at both
the medium and expanded width classes — and the gate worth re-testing:
an opt-in with no language pair must not put a chip on the screen.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
@jvsena42
jvsena42 merged commit c1937bd into main Sep 23, 2026
8 checks passed
@jvsena42
jvsena42 deleted the feat/deck-detail-study-features branch September 23, 2026 12:04
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant