A BC stream is a big-endian binary stream of instructions which describe a signed distance field.
There are four primitive types:
| ID | Name | Description |
|---|---|---|
| 0 | Void | No value. |
| 1 | Boolean | True or false. |
| 2 | Number | Decimal of undefined precision. |
| 3 | Reserved | Currently unused. |
An opcode is a 16-bit unsigned integer.
| Start Bit | End Bit | Description |
|---|---|---|
| 0 | 1 | Result primitive type. |
| 2 | 3 | Argument A primitive type. |
| 4 | 5 | Argument B primitive type. |
| 6 | 7 | Argument C primitive type. |
| 8 | 15 | Opcode ID. |
| Opcode | Result | A | B | C | ID | Description | Undefined Behavior |
|---|---|---|---|---|---|---|---|
| 5000 | Boolean | Boolean | Void | Void | 00 | Not | |
| 5400 | Boolean | Boolean | Boolean | Void | 00 | And | |
| 5401 | Boolean | Boolean | Boolean | Void | 01 | Or | |
| 5402 | Boolean | Boolean | Boolean | Void | 02 | Equal | |
| 5403 | Boolean | Boolean | Boolean | Void | 03 | Not Equal | |
| 5500 | Boolean | Boolean | Boolean | Boolean | 00 | Conditional (A ? B : C) | |
| 6800 | Boolean | Number | Number | Void | 00 | Greater Than | A = B |
| 80** | Number | Void | Void | Void | ** | Parameter | |
| 9A00 | Number | Boolean | Number | Number | 00 | Conditional (A ? B : C) | |
| A000 | Number | Number | Void | Void | 00 | Negate | |
| A001 | Number | Number | Void | Void | 01 | Sine | |
| A002 | Number | Number | Void | Void | 02 | Cosine | |
| A003 | Number | Number | Void | Void | 03 | Tangent | |
| A004 | Number | Number | Void | Void | 04 | Arc Sine | |
| A005 | Number | Number | Void | Void | 05 | Arc Cosine | |
| A006 | Number | Number | Void | Void | 06 | Arc Tangent | |
| A007 | Number | Number | Void | Void | 07 | Hyperbolic Sine | |
| A008 | Number | Number | Void | Void | 08 | Hyperbolic Cosine | |
| A009 | Number | Number | Void | Void | 09 | Hyperbolic Tangent | |
| A00A | Number | Number | Void | Void | 0A | Hyperbolic Arc Sine | |
| A00B | Number | Number | Void | Void | 0B | Hyperbolic Arc Cosine | |
| A00C | Number | Number | Void | Void | 0C | Hyperbolic Arc Tangent | |
| A00D | Number | Number | Void | Void | 0D | Absolute | |
| A00E | Number | Number | Void | Void | 0E | Square Root | |
| A00F | Number | Number | Void | Void | 0F | Floor | |
| A010 | Number | Number | Void | Void | 10 | Ceiling | |
| A011 | Number | Number | Void | Void | 11 | Natural Logarithm | |
| A012 | Number | Number | Void | Void | 12 | Logarithm 10 | |
| A013 | Number | Number | Void | Void | 13 | Natural Power | |
| A800 | Number | Number | Number | Void | 00 | Add | |
| A801 | Number | Number | Number | Void | 01 | Subtract | |
| A802 | Number | Number | Number | Void | 02 | Multiply | |
| A803 | Number | Number | Number | Void | 03 | Divide | B = 0 |
| A804 | Number | Number | Number | Void | 04 | Power | A < 0 |
| A805 | Number | Number | Number | Void | 05 | Modulo | |
| A806 | Number | Number | Number | Void | 06 | Arc Tangent | |
| A807 | Number | Number | Number | Void | 07 | Minimum | |
| A808 | Number | Number | Number | Void | 07 | Maximum |
Each instruction is its opcode, followed by each of its arguments in order. An argument starts with a u16.
| Argument | Meaning |
|---|---|
| 0x0000-0xFFFC | The result of a previous instruction |
| 0xFFFD | False |
| 0xFFFE | True |
| 0xFFFF | IEEE 32-bit float constant follows |
It is an error to use the result of the instruction for which arguments are being specified, one which will be defined later in the stream, or one which never will.
It is also an error to use an argument of a primitive type which does not match that of the parameter it is providing a value for.
The result of the last instruction in the stream is taken to be the distance to the nearest surface. It is an error to have the last instruction in the stream return any type but number.
An empty stream is considered to be a constant of infinity.
