Giving the Globe a Point of View
In God's Eye View, selecting an object does more than make it brighter. Selection can define the center of a 250-kilometer context window, change which counts appear, assign a camera owner, and open a mode with its own allowed data. This chapter follows those transitions through Contacts, aircraft tracking, Cockpit, and Space Missions. The practical judgment is important: a focused view can make relationships easier to investigate, but it must not turn proximity into threat, coverage into certainty, or a replay into telemetry.
A selected subject defines scope, not significance
The Context system begins with a subject and a radius. Around a selected aircraft, vessel, or mapped installation, God's Eye View builds cohorts of nearby flights, military flights, AIS vessels, and mapped installations. It reports counts, nearby examples, source labels, freshness, and reasons for uncertainty. The radius is 250 kilometers for the air and sea proximity cohorts. It is a window for navigation and comparison, not a radar range, a sensor footprint, a weapon envelope, or an assessment of what matters operationally.
The installation cohort has an additional boundary. Mapped installations come from viewport-bounded OpenStreetMap context, so the panel labels that result as current viewport only. It does not silently expand a local map query into a complete regional survey. A site omitted from the loaded map may be unmapped, outside the request, or unavailable from the source. The absence is not evidence that the site does not exist.
The interface also distinguishes an answer from the absence of an answer. A source that is still loading and has never produced a usable record is not presented as zero. It is unknown. A previously accepted source whose freshness has degraded can remain visible as stale. A source that cannot currently support the request is unavailable. Those states matter because “zero vessels nearby” and “the vessel source has not answered yet” lead to different decisions.
The count itself is carefully shared. The Contacts panel and the voice analyst's entity-centered “how many nearby?” question use one aircraft-proximity computation. Earlier, two paths could answer from different collections: the panel used a broad live-position set while the analyst used a smaller slice of last-fix records. Both numbers could be internally defensible and still disagree badly for the same subject. Routing both surfaces through the same engine makes their aircraft count identical by construction.
That shared computation does not make every count in the application identical. God's Eye View has several legitimate scopes. The Contacts window counts contacts in the fixed radius around its subject. An analyst query can count currently loaded records for a requested view or place. A layer can also report its loaded total. The flights layer loads according to where the camera looks, so a view-scoped loaded count may be lower than a full proximity-window count after the user dives to a new location. The interface therefore names the scope in words instead of offering a bare number.
A selected subject can also move.