Slide-mode images rendered at intrinsic ratio, top-aligned inside the
swiper chain (--pc-height:auto left .swiper-wrapper content-sized), so
the taller box showed a gap at the bottom and never center-cropped like
the image cover. The video cover had the same problem through
video-player.js (height:auto, capped at 70vh).
- aksrt_render_slider scale mode emits --pc-height/--m-height 100% so
the wrapper/slide/img chain fills the ratio box and object-fit: cover
center-crops inside it (same pattern as the .scale-height widgets)
- video-player.js renders the video at height:100% + object-fit:cover;
with an undimensioned container height:100% degrades to auto,
matching the previous behavior
- drop the dead .single-cover padding-bottom rules (every producer owns
its height inline, and the rule squeezed the absolutely positioned
slide inner) and the orphaned --single-cover-scale var in
tokens.css/compat-vars.css/main.css/InlineStyleBuilder
- rebuild public bundles; extend the single-cover-height contract
The image-mode cover reserves +50px for the title/breadcrumb overlay
(calc(scale% + 50px)), but the slide cover wrapper was given a bare
inline padding-bottom by aksrt_render_slider, permanently defeating the
.cover-slide rule, and the video cover used a bare video_scale%.
- aksrt_render_slider: add scale_extra so scale mode emits
calc(ratio% + extra); the single cover passes 50px so all three
modes share one height formula
- drop the unreachable static slide fallback and function_exists guard
(the helper is unconditionally loaded by inc/bootstrap.php), the
always-true cover_url/ep_count conditions, and the dead --scale-ratio
and inner swiper padding emitted in scale mode
- remove the dead cover CSS cluster with no producer: .single-cover-slide
(main.css/main.min.css/cover.css), .single-cover-slider family
(components/base.css), .single-cover-slide-wrap and .single-cover__*
helpers (cover.css), and initSlideCovers in misc.js
- rebuild public bundles via tools/build-public-assets.mjs
- add single-cover-height-contract test covering the shared formula
and the dead-code removals
kate522