Scroll-to-load captures only the first screen of a full-height locked page#2005

Defect: Scroll-to-load captures only the first screen of a full-height locked page that keeps its content in the body.

What: A page with a doctype that locks scrolling by giving both the root and the body the full window height with overflow hidden keeps its content inside the body, and Scroll-to-load records about one screen of it (487 pixels in the probe, against 5042 for the whole page), because the scroll step never treats the body itself as something to scroll. It was found while reviewing #1996 and left out of it, because adding the body as a scroll target changes what the signed "scroll lock lifted" field means on consent-wall pages, which is a maintainer decision.

Where it is and how to reproduce it

Where: src/main/services/backgroundRenderer.ts:157 Reproduce: Recapture a page with <!doctype html>, html, body { height: 100%; overflow: hidden } and 5000 pixels of content in the body. The scroll step picks document.scrollingElement (html in standards mode) and the descendants of body, so the body, which a script can still scroll, is never scrolled and the phase stops at the first screen. The probe that measured the 487 and 5042 figures is described in #1996's validation record, in the table of pages with and without a doctype.