From 1cfe7fa53c71b2ba595c92d1bdfc560f90242f40 Mon Sep 17 00:00:00 2001 From: Claude Date: Fri, 11 Sep 2026 19:59:53 +0300 Subject: [PATCH] Say what the day-cycle layer change does to the picture, and where the ratio was measured MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit Code review r1 found two documentation claims that the task's own contract contradicts. Both changelogs promised the picture was unchanged "to the pixel". It is not: splitting the layer moves the wall hatch by a sub-pixel, and that is exactly why four day-cycle baselines are re-taken. The entry now says what a reader can check — the outline around the plan matches pixel for pixel, the hatch inside the walls is anti-aliased a shade differently. docs/SUN.md credited Firefox with the fifteen-fold ratio. That number comes from headless Chromium through CDP tracing; the Firefox profile that opened the issue measured the same cause differently — 23 MB of texture uploads per frame and about nine frames per second. Each number now names its engine. Issue: #532 User-Visible: no --- docs/CHANGELOG.md | 6 ++++-- docs/CHANGELOG.ru.md | 5 +++-- docs/SUN.md | 10 +++++++--- 3 files changed, 14 insertions(+), 7 deletions(-) diff --git a/docs/CHANGELOG.md b/docs/CHANGELOG.md index 4010e308..546bc16e 100644 --- a/docs/CHANGELOG.md +++ b/docs/CHANGELOG.md @@ -6,8 +6,10 @@ outer outline is a blur over the whole sheet, and it lived in the same layer as the plan: any mouse movement made the engine re-run it. The outline now keeps its own layer. Measured on a demo-stand pan, rasterization fell from - fifteen times a static background to the same order as one. The picture is - unchanged, to the pixel + fifteen times a static background to the same order as one. The view is the + same: the outline around the plan matches pixel for pixel, and the hatch + inside the walls is anti-aliased a shade differently — splitting the layer + moves it by a sub-pixel ([#532](https://github.com/Matysh/houseplan-card/issues/532)). ## v1.74.0 — 2026-09-11 diff --git a/docs/CHANGELOG.ru.md b/docs/CHANGELOG.ru.md index 246dafc0..1455848a 100755 --- a/docs/CHANGELOG.ru.md +++ b/docs/CHANGELOG.ru.md @@ -12,8 +12,9 @@ это размытие по габариту всего листа, и оно жило в общем слое с планом: любое движение мыши заставляло движок прогонять его заново. Теперь контур держится отдельным слоем. Замер на панораме демо-стенда: растеризация упала с - пятнадцатикратной к статичному фону до сопоставимой с ним. Картинка - не изменилась ни на пиксель + пятнадцатикратной к статичному фону до сопоставимой с ним. Вид прежний: + ореол вокруг плана совпадает пиксель в пиксель, а штриховка внутри стен + сглаживается чуть иначе — разделение слоёв сдвигает её на субпиксель ([#532](https://github.com/Matysh/houseplan-card/issues/532)). ## v1.74.0 — 2026-09-11 diff --git a/docs/SUN.md b/docs/SUN.md index 0e8cc39d..864a4446 100644 --- a/docs/SUN.md +++ b/docs/SUN.md @@ -79,9 +79,13 @@ The public setting remains a two-value selector. `static` uses `bg_color`. is kept on its own compositing layer (`will-change: filter`, scoped to the day-cycle stage). The group holds paper silhouettes only and does not change while the plan is hovered or panned; without the hint every repaint of the - plan re-ran three blur passes over the whole sheet, which cost a Firefox - window about fifteen times the rasterization of a static background (#532). - A static background carries neither the filter nor the hint. + plan re-ran three blur passes over the whole sheet. Measured in headless + Chromium (CDP tracing, summed `RasterTask`) on a demo-stand pan, that cost + about fifteen times the rasterization of a static background; the owner's + Firefox profile that opened #532 showed the same cause as 23 MB of texture + uploads per frame and about nine frames per second, without a comparable + ratio of its own. A static background carries neither the filter nor the + hint. - Only the environment and the zero-offset alpha-aware outline outside the grouped plan-paper footprint change. The plan, paper, floors, room fills, Glow/spill, devices, labels, decor/backdrop, vacuum, hover, and window rays