Hollang

Hollang short for holographic language, is a esoteric language by User:Doesthiswork where the only way of building larger structures is to holographically embed them into variables.

Syntax
::= | ":" | "[" ("," )* "]" | "(" ")" | "!" ::= "{" * "}" ::= | "label" | "call" | "print" | "="

Explanation
Each variable is initialized to a random collection of bits these operations allow us to build associations so we can ask questions like "what is the capital of Sweden?" and "what is the Stockholm of of France?". sweden = [capital:stockholm, weather:cold] france = [capital:paris, weather:temperate] print capital:sweden print stockholm:sweden:france prints stockholm, 0.499 paris, 0.246
 * ":" xors two values
 * "[,,,]" averages the values separated by commas
 * "=" assigns the value on the right to the name on the left
 * print prints the name of the variable closest in value to the value on its right, followed by how close of a match it is.


 * label associates a value with a piece of code so that you can later execute that code
 * call finds the code that is labeled by the value to its right and executes it
 * "( ... )" coerces the value inside to become the closest named value. This partial matches to be sharpened to be a full match
 * "!:" inverts the value to its right allowing you to avoid associating a name with it.

Example
label loop { print [blue, red] counter = [counter, [counter, [counter ,end]]] call counter } label end { print end } counter = loop call loop prints blue, 0.511 red, 0.506 red, 0.504 red, 0.502 blue, 0.511 red, 0.508 end, 1

Addition
A way to add 2 number mod 10, the !: allows one to use the opposite of a value to keep from giving a name to the value. odd = [1,3,5,7,9] even = [0,2,4,6,8] 01 = [0:1, 2:3, 4:5, 6:7, 8:9] 03 = [1:2, 3:4, 5:6, 7:8, 9:0]  inc = [odd:03, even:01] dec = [odd:01, even:03] label step { a=!: (a:!:(dec:a:!)) b=!: (b:!:(inc:b:!)) call [0:0, 0:0, a:!:step]:a:! } label 0 {print b:!} a=!: 7 b=!: 2 call step produces ready 9, 1

Computational Class
Hollang is, strictly speaking, a finite-state automaton, bounded by word size much like the Von-Neumann architecture ones on which it runs; extending it to employ infinite random bitvectors, however, is considered trivial, and left as an excercise for the reader.

More tractably, the interpreter could track the maximum number of (inherently destructive) averagings performed on each value, and rerun the program with increased representation size when this starts to violate recoverability. Assuming memory than can guarantee recognition of mathematical proximity after arbitrary dilution, Church Numerals represented homeopathically as  [!:0 [!:0 [!:0 [!:0 1]]]] can then be used in a standard minsky machine-based completeness proof: each stack represented as a pair of incrementable "pop" and "push" registers, whose equality signals underflow.

External resources
Hollang html file