One attempt at an answer: a coordinate reference system whose origin, reference meridian and direction are read from the sun rather than chosen by a nation — and, in its SPPS-25 form, every point on Earth expressed as a single sub-meter-accurate number.
Twice a year, a solstice marks the moment when the South Pole is most tilted toward the sun and most tilted away. And once a year, Earth reaches perihelion — its closest point to the sun in its elliptical orbit.
These events drift slowly relative to one another, because the calendar year and Earth's orbital year aren't quite the same length. Roughly every 21,000 years, that drift completes a full cycle — and passes through a moment where the December solstice and perihelion land at exactly the same instant.
The most recent such moment is well documented: December 1246 CE. At that instant, the South Pole was tilted at its most extreme toward the sun (southern midsummer) at precisely the same time Earth's orbit brought it physically closest to the sun. Both effects point the same direction and stack.
This isn't a symbolic description. It's the literal physical consequence of two unrelated astronomical cycles briefly agreeing with each other — reproducible by anyone with orbital mechanics and no observatory required. This moment of greatest proximity is what the system is named for.
At the instant of the 1246 solstice/perihelion coincidence, the sun stood directly over one meridian — the line along which it was local apparent noon. That line is the SPPS reference meridian.
Seen from the South Pole itself, where every direction is north, it is simply the direction of the sun at that moment. No survey, observatory or vote is needed to find it.
Why noon? Sunrise and sunset each carry cultural weight: some traditions begin the day at dawn, others at dusk. Noon — the sun at its highest — favors neither, and it ties the meridian to the very same instant that defines the system, rather than to a separate event.
Computing it requires care: the event's exact time has to be converted from a uniform time scale into Earth's actual historical rotation state, using measured ΔT from the eclipse-record literature (Morrison & Stephenson, 2004) — not assumed to be zero, the way a casual calculation might.
In its simplest form, SPPS is an ordinary two-number coordinate pair, re-anchored: distance from the South Pole, and angle east of the SPPS meridian.
Because neither value is ever negative, there are no hemisphere suffixes to drop or misread. And because an r above 90 or a λ above 180 can't be an ordinary latitude or longitude, most SPPS coordinates identify their own system even when stripped of metadata.
Converting to or from WGS84 is a one-line transform. This is the adoptable version of the idea; the next step is the more radical one.
To collapse the pair into a single number, tie the angle to the distance from the Pole through a spiral that winds from the South Pole to the North Pole. A location's position along that one spiral becomes its entire address.
That many windings puts adjacent turns just 1.19 m apart, while staying comfortably inside what a standard 64-bit float can represent. The address is one decimal: 0.0 at the South Pole, 0.5 where the Equator crosses the SPPS meridian, 1.0 at the North Pole.
A spiral that fine can't be drawn on a globe. What can be shown is what each digit of an address actually does: the first seven pick out a ring around the whole Earth, and the last nine slide along it.
Or click anywhere on the map below to get an address directly, checked against the actual click rather than just claimed.
Every country boundary on Earth — 177 countries, 289 rings, 10,654 vertices, from Natural Earth's public dataset — converted into SPPS-25 addresses and back, rendered with no basemap underneath: nothing on screen except lines decoded from single 16-digit numbers, live, in JavaScript.