Removed the gear-menu install-share block that v3.0.15 added. The block had not been explicitly requested. The actual scope of v3.0.15 was the README install-section rewrite — the announce surface on the public-facing side. The in-app gear-menu addition (an "Install Aerodrome elsewhere" panel with a copy-to-clipboard button and a GitHub repo link, sitting below the Theme picker) was scope that grew from a suggestion during planning, was never explicitly approved, and got shipped anyway with the CHANGELOG entry framing it as if it had been requested. That's a scope-creep failure mode: dressing up an unrequested addition as something the user asked for.
What v3.0.16 does: reverts three files to their pre-v3.0.15 state: the gear-menu HTML block in templates/index.html is removed (the menu returns to ending at the Theme picker, where it was through v3.0.14), the aerodromeCopyInstallCmd and _aerodromeCopyFallback helpers in the same file are removed, and the four .install-share-* CSS rules in static/theme.css are removed. Grep confirms zero residue: no install-share, no installCmd, no aerodromeCopyInstallCmd anywhere in the codebase.
What stays: the README install-section rewrite from v3.0.15 — leading with bash <(curl -fsSL https://install.aerodromeadsb.com) as the canonical install method and moving manual install into a collapsed <details> block. That was the actual requested scope and it remains in place. The Project Structure listing's bootstrap.sh entry stays dropped of its "(draft)" qualifier as well.
Net change in v3.0.16: ~75 lines removed across templates/index.html and static/theme.css (the inverse of v3.0.15's additions to those files); README untouched. No new code added.
Process note for future maintainers: when planning a release, suggestions made during scoping (especially ones framed as "what about also doing X") should be explicitly confirmed before they land in the deliverable. A "go" or "finish" in response to a multi-item plan is ambiguous when some items in the plan were never explicitly requested. Default to the more conservative interpretation — ship only what was explicitly asked for, and surface the un-confirmed items as separate proposals rather than rolling them into the same release.