. That wrapper sits directly * inside

, and app.css:806 declares * * .hbox > aside, .hbox > section { display: table-cell; height: 100%; } * * This is SELECTOR matching, not box inheritance. Before the wrapper, the * view's own

was the direct child of .hbox, so it * matched that rule and received height:100%. After the wrapper, the WRAPPER is * the direct child, so the wrapper takes the table-cell slot - and the view's * section, now a grandchild, matches nothing and falls back to a plain block * with height:auto. Inside a table-cell with no intrinsic height that resolves * to 0. app.css:781 sets `body { overflow: hidden }`, so everything below the * first ~50px was clipped away. The content was in the DOM the whole time - the * form measured 1204px tall - it just could not paint. * * Six views render through _CustomerLayout with

as * their root and were affected: Customer/Settings, Customer/CustomerHome, * Customer/Project, Customer/ProjectMeetingSeriesAdmin, Audience/AudiencesHome, * Taxonomy/Index. Views rooted in .vbox set their own height:100% and were not. * * WHAT THE ORIGINAL VERIFICATION ACTUALLY COVERED - READ THIS BEFORE TRUSTING IT * ----------------------------------------------------------------------------- * The comment this replaces claimed the geometry was measured before and after * and was unchanged. That claim was WRONG, and it survived three review rounds. * * Measured: WIDTH and horizontal position, on /Customer/Settings only * (x=200, w=1080 before and after). Both were genuinely * unchanged - a table-cell fills the row's width either way. * NOT measured: HEIGHT. Height is the axis the wrapper broke, and nothing in * the round-1 validation looked at it. * NOT measured: any _CustomerLayout view other than /Customer/Settings. * * Nor did the scanner catch it: axe skips descendants of a zero-size clipped * ancestor, so a page whose entire content section collapsed to height 0 * reports FEWER findings, not more. The clean axe runs on these pages after * round 1 were false passes. * * THE FIX * ------- * Give the inner section back the height the selector no longer grants it. * display:contents on the wrapper is NOT a substitute: it removes the wrapper's * box but does nothing to selector matching, so the inner section still fails * to match `.hbox > section` and still never receives height:100%. * * The same interception applies to bootstrap-3-shim.css:1157 (`display: * table-cell !important` on .hbox > section) and to the unscoped duplicate of * the app.css rule in Civic-Front.css:2731. Both only set display, which the * wrapper now legitimately carries; height is the only property that has to be * handed back down. * * E2E/a11y/verify-skiplink.mjs asserts the landmark is reachable. */ #legacy-main-content > section.content { height: 100%; } /* Icon-only control rendered inline in a run of text or icons. Exists so an activatable control can be a