Base-1024 Hybrid Coordinate System
Accuracy: ~20m Global
Grid: 1024^4 (1.1 Trillion)
8 characters encoding exactly 40 bits of coordinate data.
The FourWords system uses a Base-1024 recursive grid to pinpoint any roughly 20-meter square on Earth using exactly four simple English words. It acts as a direct bridge between a highly compact 40-bit machine-readable hex structure and a human-friendly naming convention.
Based on the mechanics of the Base-1024 recursive grid system, the following ASCII characters best represent a unit or block of the map in software and text output:
§ (ASCII 167) — The most accurate visual representation. Since the system divides the Earth into 1024^4 distinct spatial sections, this character perfectly represents the literal fractal nature of the standard (§blue.moon.fast.jump) is exactly identical to the 40-bit machine string (§F9X7A23B).The FourWords Base-1024 system, including its hybrid machine/human-readable grid logic, mathematical algorithms, and the 1024-word dictionary, is built to be a free to use and replicate system.
This specification is released into the Public Domain (CC0 / The Unlicense). It is completely free to use, modify, distribute, and commercialize.
There are no proprietary API locks, no subscription fees, and no patent restrictions. Anyone can adopt this system to build localized mapping tools, offline rescue beacons, drone navigation protocols, or mesh networks.
The FourWords system offers a revolutionary approach to global mapping by acting as a flawless bridge between machine-level efficiency and human-friendly communication.
blue.moon.fast.jump is in the same neighborhood as blue.moon.fast.stay.Because traditional computing architecture handles data in 64-bit integers (like database BIGINT fields), and FourWords natively only requires 40 bits, the system leaves exactly 24 bits free per coordinate. This unlocks massive systemic optimizations without requiring additional database columns:
A known limitation of standard recursive grids (like Quadtrees or Geohashes) is the "boundary problem." In traditional linear mappings (-180 to 180), crossing the Equator or Prime Meridian causes every single bit to flip, completely changing all four words.
The Rectification: Sign-Magnitude Geometry
By mathematically mapping the grid Center-Out, FourWords isolates major boundary crossings. If you step across the Equator, only the 20th bit (Hemisphere) flips. This means Word 1 changes, but Words 2, 3, and 4 remain identical because they radiate outward symmetrically from the origin.
For programmatic proximity matching across minor boundaries, developers simply convert the target 4-word address back into its raw latInt and lngInt 20-bit coordinates, add/subtract 1 to find adjacent blocks, and generate a "3x3 Buffer" mathematically without needing heavy lookup tables.
The core of FourWords is its seamless, offline translation between standard WGS84 GPS coordinates, a 40-bit Base-32 machine string, and the 4-word text string.
To prevent catastrophic boundary flips at the Equator and Prime Meridian, the Earth is mapped using a "center-out" sign-magnitude integer space (2^20 = 1,048,576).
To achieve the recursive "zoom" effect, we interleave the coordinates. We take 5 bits of latInt and 5 bits of lngInt to create a 10-bit integer (giving a value from 0 to 1023).
Because each word represents exactly 10 bits of data, it can be perfectly split into two 5-bit chunks. Each 5-bit chunk (0-31) maps exactly to a Crockford-style Base-32 character (0123456789ABCDEFGHJKMNPQRSTVWXYZ).
1 Word = 10 bits = Two Base-32 Characters. The entire 4-word coordinate thus equates to a flawless 8-character Base-32 string.
Because the bits are interleaved from most significant to least significant, the words naturally represent bounding boxes of increasing accuracy. If two coordinates share the same first words, they are physically close to each other.
| Level | Word Count | Bit Depth | Approx. Area Size | Geographic Scale |
|---|---|---|---|---|
| Level 1 | 1 Word | 10 bits | ~600 km x 600 km | Quadrant / Continent |
| Level 2 | 2 Words | 20 bits | ~18 km x 18 km | Region / City Bounds |
| Level 3 | 3 Words | 30 bits | ~600 m x 600 m | Neighborhood / District |
| Level 4 | 4 Words | 40 bits | ~19 m x 19 m | Specific Plot / Building |
The official dictionary is strictly 1024 words long. To maximize human readability, it prioritizes simple, distinct, "Grade 1" 4-letter English nouns and verbs, explicitly excluding homophones (blue/blew), offensive terminology, and visually confusing letters.
The foundational list consists of hyper-common, universally recognized English words (visible in the final list below).
To hit the exact mathematical requirement of 1024 without relying on obscure and unpronounceable words, the remaining 668 slots are generated programmatically using a deterministic CVCV (Consonant-Vowel-Consonant-Vowel) Generator.
The generator iterates the first consonant fastest (e.g., baba, daba, faba) to guarantee an even, alphabetical distribution and prevent a cluster of words starting with the same letter. To maximize auditory clarity over radio or phone, phonetically ambiguous or globally inconsistent consonants (c, h, j, q, v, w, x, y, z) are explicitly omitted. This leaves only 12 ultra-clear, hard-sounding consonants. This creates short, distinct, highly pronounceable phonetic blocks universally readable across global languages. By hardcoding the generation algorithm rather than the string list, offline applications can dynamically construct the exact 1024-word array at runtime.
This is the final, deterministic array of words generated by the combination of the seed list and the distributed CVCV algorithm:
Loading...