Cache generation stale-write guard
A generation pattern that blocks late-finishing requests from re-writing cache when async deletion overlaps stale-while-revalidate
Knowledge documents in Browser.
A generation pattern that blocks late-finishing requests from re-writing cache when async deletion overlaps stale-while-revalidate
A pattern to stabilize Canvas 2D graph node color and twinkle via ID hash, and restore globalAlpha and shadow state
SWR pattern in Service Worker Cache Storage: return cache immediately and refresh in the background, with capacity and security caveats
A pattern for building scroll-entry fade/slide with animation-timeline view() and animation-range entry, without JS
A typography strategy that uses the optional display to prevent late webfont swap and reduce tag/body layout shift (CLS)
pointercancel, touch-action, and coarse-pointer hover fallbacks to stabilize custom drag·resize on mobile
React commit attaches child DOM first, so an ancestor ref may be null at the time of a child useLayoutEffect — defer ancestor DOM side effects to useEffect
Why a popover/toolbar pinned with coordinates measured only once to a position:fixed anchor drifts off the text after scrolling, and the scroll/resize re-measurement pattern
Why a text-selection-based floating toolbar is driven by selectionchange instead of a mousedown/mouseup gate, and the ownership pattern for hybrid editors
A pattern for explicitly delegating to document undo via chainCommands when a rich-text engine preventDefaults Mod-Z/Y during focus
The principle of preventing composition breakage in a CJK IME composition session with contenteditable/React editors via commit defer and mount-once
A browser-native drag session API that implements moves crossing subtree boundaries via the OS drag loop and DataTransfer
user-select:none only blocks drag selection and does not fully exclude text from the Selection/clipboard; to exclude it, CSS generated content is safer
Why CLS occurs when a loading skeleton occupies different space than the final UI, and the height/width reservation prescription to prevent it
An experiment design that avoids dev-server trace outliers and compares before/after with an N-run median
How to distinguish that the Lighthouse Performance score, category audits, and Chrome trace ms are each different instrumentations
The stages a browser goes through from receiving HTML/CSS/JS to painting the first screen pixels, and how to shorten that path