Skip to content

Latest commit

 

History

History
2531 lines (1945 loc) · 61.3 KB

manual.md

File metadata and controls

2531 lines (1945 loc) · 61.3 KB

% FastBasic %VERSION% - Fast BASIC interpreter for the Atari 8-bit computers

Introduction

FastBasic is a fast interpreter for the BASIC language on the Atari 8-bit computers.

One big difference from other BASIC interpreters in 1980s era 8-bit computers is the lack of line numbers, as well as an integrated full-screen editor. This is similar to newer programming environments, giving the programmer a higher degree of flexibility.

Another big difference is that default variables and operations are done using integer numbers; this is one of the reasons that the programs run so fast relative to its peers from the 1980s.

The other reason is that the program is parsed on run, generating optimized code for very fast execution.

Currently, FastBasic supports:

  • Integer and floating point variables, including all standard arithmetic operators.
  • All graphic, sound, and color commands from Atari BASIC, plus some extensions from TurboBASIC XL.
  • All control flow structures from Atari BASIC and TurboBASIC XL.
  • Automatic string variables of up to 255 characters.
  • Arrays of "word", "byte", floating point and strings.
  • User defined procedures, with integer parameters.
  • Compilation to binary loadable files.
  • Available as a full version FB.COM, as a smaller integer-only FBI.COM, and as a command-line compiler FBC.COM and FBCI.COM.

First Steps

To run FastBasic from the included disk image, simply type FB at the DOS prompt. This will load the IDE and present you with a little help text:

--D:HELP.TXT-------------------0--
'  FastBasic %VERSION% - (c) 2025 dmsc
'
' Editor Help
' -----------
'  Ctrl-A : Move to beginning of line
'  Ctrl-E : Move to end of line
'  Ctrl-U / Ctrl-I : Page up / down
'  Ctrl-Z : Undo (only current line)
'  Ctrl-C : Set Mark to current line
'  Ctrl-V : Paste from Mark to here
'  Ctrl-Q : Exit to DOS
'  Ctrl-S : Save file
'  Ctrl-L : Load file
'  Ctrl-N : New file
'  Ctrl-R : Parse and run program
'  Ctrl-W : Compile to binary file
'  Ctrl-G : Go to line number
'
'- Press CONTROL-N to begin -

You are now in the integrated editor. On the first line of the screen the name of the currently edited file is shown, and at the right the line on which the cursor is located. Please note that lines that show an arrow pointing to the top-left are empty lines beyond the last line of the current file.

In this example, the cursor is in the first column of the first line of the file being edited.

As the help text says, just press the CONTROL key and the letter N (at the same time) to begin editing a new file. If the text was changed, the editor asks if you want to save the current file to disk; to skip saving simply type CONTROL-C; to cancel the New File command type ESC.

Now you are ready to start writing your own BASIC program. Try the following example, pressing RETURN after each line to advance to the next:

INPUT "WHAT IS YOUR NAME?";NAME$
?
? "HELLO", NAME$

The parser will let you know if you made any mistakes. To make corrections move back using the cursor keys, this is CONTROL and -, =, + or *, then press BACKSPACE key to delete the character before the cursor or press DELETE (CONTROL and BACKSPACE) to delete the character below the cursor. To join two lines, go to the end of the first line and press DELETE.

After typing the last line, you can run the program by pressing CONTROL and R.

If there are no errors with your program, it will be run now: the computer screen will show WHAT IS YOUR NAME?, type anything and press RETURN, the computer will reply with a greeting and the program will end.

After this, the IDE waits for any key press before returning to the editor, so you have a chance to see your program's output.

If you press the BREAK key when the program is running, it will terminate, wait for a key press and return to the IDE.

If you made a mistake typing in the program code, instead of the program running, the cursor will move to the line and column of the error so you can correct it and retry.

Remember to save often by pressing the CONTROL and S keys and entering a filename. Type the name and press ENTER to save. As with any prompt, you can press ESC to cancel the save operation. Use the BACKSPACE over the proposed file name if you want to change it.

Compiling The Program To Disk

Once you are satisfied with your program, you can compile to a disk file, producing a program that can be run directly from DOS.

Press the CONTROL and W key and type a filename for your compiled program. It is common practice to name your compiled programs with an extension of .COM or .XEX. With .COM extension files you don't need to type the extension in some versions of DOS on the Atari. The .XEX name is common in modern times to distinguish Atari executables from MSDOS/Windows programs (which are usually .EXE or sometimes .COM)

Compiled programs include the full FastBasic runtime, so you can distribute them alone without the IDE.

You can also compile a program directly from DOS by using the included command line compiler FBC.COM. You can provide the BASIC source and the compiled output file names in the command line; if not given the compiler will prompt you to input a file name.

If you don't use floating point, using the integer versions (FBI and FBCI) will compile to a smaller file.

Advanced Editor Usage

The editor includes a few commands, most of those are already explained above.

  • CONTROL-A and CONTROL-E Moves the cursor ro the beginning or the end of the line respectively.

  • CONTROL-U and CONTROL-I Moves the cursor 19 lines up or down respectively.

  • CONTROL-G Moves the cursor to a specific line.

  • CONTROL-Z Reverts all editing of the current line. Note that changing the line clears the undo buffer, so you can't undo more than one line.

  • CONTROL-C Sets the current line as the source for a copy operation.

  • CONTROL-V Copy one line from the source marked with the CONTROL-C to the current cursor position. After the copy, the source line is advanced, so by pressing CONTROL-V multiple times you can copy multiple consecutive lines.

  • CONTROL-L and CONTROL-S Loads a file, or Saves the file being edited, respectively.

  • CONTROL-Q Returns to DOS, abandoning the changes in the current file.

  • CONTROL-R Parses the current program and runs it.

  • CONTROL-W Compiles the current program and saves it to a binary file.

Making the Editor Faster

The FastBasic IDE uses the Atari screen handler for writing text, so it is compatible with all 80 column and other expansions available.

As the original screen handler in the Atari OS is slow, there is a screen accelerator included on the FastBasic disk.

To use the accelerator, just type EFAST in the DOS prompt, before loading the IDE, and enjoy editing programs faster.

About The Syntax

The syntax of FastBasic language is similar to many BASIC dialects, with the following main rules:

  1. A program line must be of 4 types:

    • a comment line, starting with a dot . or an apostrophe ',
    • a statement followed by its parameters,
    • a variable assignment, this is a name followed by = and the expression for the new value. For string variables, there is also a concatenation operator, =+.
    • an empty line.

  2. All statements and variable names can be lower or uppercase, the language is case insensitive.

  3. Statements can be abbreviated to reduce typing; each statement has a different abbreviation.

  4. Multiple statements can be put on the same line by placing a colon : between statements.

  5. After any statement a comment can be included by starting it with an apostrophe '.

  6. No line numbers are allowed.

  7. Spaces after statements and between operators are optional and ignored.

In the following chapters, whenever a value can take any numeric expression, it is written as "value", and whenever you can use a string it is written as "text".

Expressions

Expressions are used to perform calculations in the language.

There are numeric expressions (integer and floating point), boolean expressions, and string expressions.

In FastBasic, standard numeric expressions are evaluated as integers from -32768 to 32767, this is called 16 bit signed integer values.

Floating point expressions are used only if numbers have a decimal point. Floating point numbers are stored with standard Atari binary-coded decimal (BCD) representation, with a range from 1E-98 to 1E+98.

Boolean expressions are "true" or "false", represented as the numbers 1 and 0, respectively.

String expressions contain arbitrary text. In FastBasic strings can have up to 255 characters of length. (This is similar to Microsoft BASICs found on other 8-bit microcomputers, and in contrast to the 32K limit found in Atari BASIC and TurboBASIC XL.)

Numeric Values

Integer values can be written as decimal numbers (like 123, 1000, etc.), as hexadecimal numbers with a $ sign before (like $1C0, $A00, etc.) or by using the name of a variable.

Floating point values are written with a decimal dot and an optional exponent (like 3.14159, -3.2, or 1.0E+10).

Numeric Variables

Variable names must begin with a letter or the symbol _, and can contain any letter, number or the symbol _. Examples of valid integer variable names are COUNTER, My_Var, num1.

Floating point variables have an % as last character in the name. Examples of valid floating point variable names are MyNum%, x1%.

In FastBasic, variables can't be used in an expression before being assigned a value; the first assignment declares the new variable.

Numeric Operators

There are various "operators" that perform calculations in expressions. The operators with higher precedence always executed first.

The order of precedence for integer operators are:

  • + - : addition, subtraction, from left to right.
  • * / MOD: multiplication, division, modulus, from left to right.
  • & ! EXOR: binary 'and', 'or', 'exclusive or', from left to right.
  • + - : positive / negative.

For example, an expression like: 1 + 2 * 3 - 4 * -5 is evaluated in the following order:

  • First, the unary - before the 5, giving the number -5.
  • The first multiplication, giving 2*3 = 6.
  • The second multiplication, giving 4*-5 = -20.
  • The addition, giving 1+6 = 7.
  • The subtraction, giving 7 - -20 = 27.

So, in this example the result is 27.

If there is a need to alter the precedence, you can place expressions within parenthesis (e.g., "(2+5)*10", which results in 70).

Note that MOD and EXOR can be abbreviated M. and E. respectively.

When using floating point expressions, the operators are:

  • + - : addition, subtraction, from left to right.
  • * / : multiplication, division, from left to right.
  • ^ : exponentiation, from left to right.
  • + - : positive / negative.

Note that integer expressions are automatically converted to floating point if needed, as this allows mixing integers and floating point in some calculations, but you must take care to force floating point calculations to avoid integer overflows.

Example: the expression -

a% = 1000 * 1000 + 1.2

gives correct result as 1000 is converted to floating point before calculation, but:

x=1000: a% = x * x + 1.2

gives incorrect results as the multiplication result is bigger than 32767.

Note that after any floating point errors (division by 0 and overflow), ERR() returns 3.

Boolean Operators

Boolean operators return a "true" or "false" value instead of a numeric value, useful as conditions in loops and IF statements.

Note that any boolean value can also be used as a numeric value, in this case, "true" is converted to 1 and "false" converted to 0.

The supported boolean operators, in order of precedence, are:

  • OR : Logical OR, true if one or both operands are true.
  • AND : Logical AND, true only if both operands are true.
  • NOT : Logical NOT, true only if operand is false.
  • <= >= <> < > = For integer or floating point comparisons, compare the two numbers and return true or false. Note that <> is not equal. You can only compare two values of the same type, so an expression like x = 1.2 is invalid, but 1.2 = x is valid as the second operand is converted to floating point before comparison.

The words OR, AND and NOT can be abbreviated O., A. and N.

Arrays

Arrays hold many ordered values (called elements). The array elements can be accessed by an index.

In FastBasic, arrays must be dimensioned before use (see DIM statement below). The index of the element is written between parentheses and goes from 0 to the number of elements. Note that FastBasic does not check for out of boundary accesses, so you must be careful with your code to not overrun the size of the arrays.

You can use an array position (the variable name followed by the index) in any location where a standard numeric variable or value is expected.

Arrays can be of four types:

  • WORD arrays (the default if no type is given) use two bytes of memory for each element, and works like normal numeric integer variables: -32768 to 32767 (signed).
  • BYTE arrays use only one byte for each element, so the numeric range is reduced from 0 to 255 (unsigned).
  • Floating point arrays, works like any floating point variable, and use six bytes of memory for each element.
  • String arrays store a string in each element. String arrays use two bytes of memory for each element that is not yet assigned (containing empty strings), and 258 bytes for each element with a string assigned.

String Values

String values are written as a text surrounded by double quotes ("). If you need to include a double quote character in a string, you must write two double quotes together.

Example:

PRINT "Hello ""world"""

Will print:

Hello "world"

You can also include any character with it's hexadecimal code using $ just after the closing quote, with no spaces around. This is the only way to include an ENTER character inside a string constant, see this example:

PRINT "Hello"$9B"world"$2E$2E

Will print:

Hello
world..

The bracket operator [ ] allows creating a string from a portion of another, and accepts two forms:

  • [ start ] This form selects all characters from start up to the end of the string, counting from 1. So, A$[1] selects the entire string, while A$[3] selects from the third character to the end, effectively removing the two leftmost characters.

  • [ start , len ] This form selects at most len characters from start, or up to the end of the string if there is not enough characters.

Example:

PRINT "Hello World"[7]
A$ = STR$(3.1415)[3,3]
? A$
? A$[2,1]

Will print:

World
141
4

Note that the bracket operator creates a new string and copies the characters from the original string to the new one. As the same buffer is always used for the new string, you can't compare two values without first assigning one of them to a new string variable.

This will print ERROR:

A$="Don't Work"
IF A$[2,2] = A$[3,3] THEN ? "ERROR"

While this will print GOOD:

A$="Long string"
B$=A$[2,2]
IF B$ <> A$[3,3] THEN ? "GOOD"

String Variables

The naming convention for string variables is the same as for numeric variables, but must end with a $ symbol. Examples of valid string variable names are Text$, NAME1$.

String variables always use 256 bytes, the first byte storing the string length and the following bytes storing up to 255 ATASCII characters.

There are two types of string assignments:

  • The standard = sign copies the string expression in the right to the variable in the left.

  • The =+ sign copies the string expression at the right to the end of the current string, concatenating the text.

Example:

A$ = "Hello "
A$ =+ "World"
? A$

Will print:

Hello World

Functions

Functions take parameters (normally between parentheses) and return a result. Functions can be abbreviated by using a shorter name ended in a dot, for example you can write R.(10) instead of RAND(10).

You can also omit parentheses on functions that take only one argument, for example, RAND 10.

Note: This is not possible when the function accepts a variable number of arguments (as with USR), or with the ADR function.

Some functions don't take parameters, and you must provide a set of parentheses, like KEY(). However, when abbreviated, you can omit the parenthesis, like K. for KEY().

Standard Functions

Following is a list of all the general purpose functions supported by FastBasic. Shown are the full syntax and the abbreviated syntax.

  • TIME / T. : Returns the current time in "jiffies." This is about 60 times per second in NTSC systems or 50 times per second in PAL systems. Use TIMER statement to reset to 0. Remember that this function returns an integer, so the maximum value is 32767, or about 9 minutes in NTSC, and a little less than 11 minutes in PAL, after this the value will become negative. If you need to measure more than this amount, consider using the floating-point version %TIME Note: TIME is special, and does not need parentheses.

  • ABS(num) / A.(num) : Returns the absolute value of num (e.g., ABS(5) and ABS(-5) both result in 5). Can be used with integers and floating point.

  • SGN(num) / SG.(num) ; Returns the sign of num, this is 1 if positive, -1 if negative or 0 if num is 0. Can be used with integers and floating point.

  • RAND(num) / R.(num) : Returns a random, non negative number, a maximum of 1 less than num. (e.g., RAND(3) will result in 0, 1, or 2.) (See also RND().)

  • FRE() / F. : Returns the free memory available in bytes.

  • ERR() / E. : Returns the last Input/Output error value, or 1 if no error was registered.

  • LEN(string) / L.(string) : Returns the length of the string.

  • VAL(string) / V.(string) : Converts string to a number. If no conversion is possible, ERR() is set to 18. Can be used with integers and floating point.

  • ASC(string) / AS.(string) : Returns the ATASCII code of the first character of the string.

Atari Specific Functions

The following functions allow interacting with the Atari hardware to read controller and keyboard input and to program with Player/Missile graphics.

  • PADDLE(n) / PA.(n) : Returns the value of the PADDLE controller n.

  • PMADR(n) / PM.(n) : Returns the address of the data for Player n or the address of the Missiles with n = -1.

  • PTRIG(n) / PT.(n) : Returns 0 if the PADDLE controller n button is pressed, 1 otherwise.

  • STICK(n) / S.(n) : Returns the JOYSTICK controller n position. STICK(_n_) values are:

    `10`  `14`  ` 6`

`11`  `15`  ` 7`

` 9`  `13`  ` 5`

  • STRIG(n) / STR.(n) : Returns 0 if JOYSTICK controller n button is pressed, 1 otherwise.

  • KEY() / K. : Returns 0 if no key was pressed, or a key code. The returned value only goes to 0 after reading the key in the OS (via a GET or POKE 764, 255 statement).

    Hint: The value returned is actually the same as (PEEK(764) EXOR 255). The following program will show the KEY() codes for pressed keys:

    PRINT "Press keys, exit with ESC"
REPEAT
  REPEAT : UNTIL KEY()
  PRINT "Key code: "; KEY()
  GET K
  PRINT "ATASCI code: "; K
UNTIL K=27

Floating Point Functions

These functions use floating point values, and are only available in the floating point version.

In case of errors (such as logarithm or square root of negative numbers and overflow in the results), the functions will return an invalid value, and the ERR() function returns 3.

  • ATN(n) / AT.(n) : Arc-Tangent of n.

  • COS(n) / CO.(n) : Cosine of n.

  • EXP(n) : Natural exponentiation.

  • EXP10(n) / EX.(n) : Returns ten raised to n.

  • INT(num) / I.(num) : Converts the floating point number num to the nearest integer from -32768 to 32767.

  • LOG(n) : Natural logarithm of n.

  • LOG10(n) / LO.(n) : Decimal logarithm of n.

  • RND() / RN. : Returns a random positive number strictly less than 1. (See also RAND().)

  • SIN(n) / SI.(n) : Sine of n.

  • SQR(n) / SQ.(n) : Square root of n.

  • %TIME / %T. : This is the same as the TIME integer function, but returning a 24 bit number that does not wrap until more than 3 days. Note: Don't use the TIMER statement if you are using this function, as the returned value will be invalid.

String Functions

  • STR$(num) : Returns a string with a printable value for num. Can be used with integers and floating point. Note that this function can't be used at both sides of a comparison, as the resulting string is overwritten each time it is called.

  • CHR$(num) : Converts num to a one character string with the ATASCII value.

Low level Functions

The following functions are called "low level" because they interact directly with the hardware. Use with care!

  • ADR(arr) / &arr : Returns the address of the first element of arr in memory. Following elements of the array occupy adjacent memory locations. Instead of ADR(X) you can simply type &X.

  • ADR(str) / &str : Returns the address of the string in memory. The first memory location contains the length of the string, and following locations contain the string characters. (This differs from Atari BASIC and TurboBASIC XL, where the address returned points to the first character of the string.)

  • ADR(var) / &var : Returns the address of the variable in memory.

  • DPEEK(addr) / D.(addr) : Returns the value of memory location addr and addr+1 as a 16 bit integer. This is the same as doing PEEK(addr)+PEEK(addr+1)*256

  • PEEK(address) / P.(address) : Returns the value of memory location at address.

  • USR(address[,num1 ...]): Low level function that calls the user supplied machine code subroutine at address.

    Parameters are pushed to the CPU stack, with the LOW part pushed first, so the first PLA returns the HIGH part of the last parameter, and so on.

    The value of A and X registers is used as a return value of the function, with A the low part and X the high part.

    This is a sample usage code snippet:

    ' PLA / EOR $FF / TAX / PLA / EOR $FF / RTS
DATA ml() byte = $68,$49,$FF,$AA,$68,$49,$FF,$60
FOR i=0 TO 1000 STEP 100
  ? i, USR(ADR(ml),i)
NEXT i

  • $(addr) : Returns the string at memory address addr.

    This is the inverse of ADR(), and can be used to create arbitrary strings in memory. For example, the following code prints AB:

    DATA x() byte = 2, $41, $42
? $( ADR(x) )

    Also, you can store string addresses to reuse later, using less memory than copying the full string:

    x = ADR("Hello")
? $( x )

  • %(n) : This returns the floating-point value stored at memory address n.

List Of Statements

In the following descriptions, statement usage is presented and the abbreviation is given after a /.

Console Print and Input Statements

Reads Key From Keyboard
GET var / GE.

Waits for a key press and writes the key value to var, which can be a variable name or an array position (like "array(123)").

Note: Some keys on the Atari -- the console keys START, SELECT, and OPTION; modifiers SHIFT and CONTROL; and the HELP, BREAK, and RESET keys -- are not handled in the same way as the main keyboard, and cannot be read by GET.

Hints: The status of all three console keys may be read via the GTIA CONSOL register, PEEK(53279).

Whether the HELP key is pressed can be detected via the POKEY KBCODE register, PEEK(53769).

Whether either SHIFT key is pressed can be detected via the POKEY SKCTL register, PEEK(53775).

Input Variable Or String
INPUT var / I.
INPUT "prompt"; var
INPUT "prompt", var INPUT ; var

Reads from keyboard/screen and stores the value in var.

A "?" sign is printed to the screen before input, or the "prompt" if given. Also, if there is a comma after the prompt, spaces are printed to align to a column multiple of 10 (similar to how a comma works in PRINT). In the case you don't want any prompt, you can use a semicolon alone.

If the value can't be read because input errors, the error is stored in ERR(). Valid errors are 128 if BREAK key is pressed and 136 if CONTROL-3 is pressed.

In case of a numeric variable, if the value can't be converted to a number, the value 18 is stored in ERR().

See the Device Input and Output Statements section for the INPUT # usage.

Moves The Screen Cursor
POSITION column, row / POS.

Moves the screen cursor position to the given column and row, so the next PRINT statement outputs at that position.

Rows and columns are numerated from 0.

Print Strings And Numbers
PRINT expr, ... / ?
PRINT expr TAB(expr) ...
PRINT RTAB(expr) ...
PRINT COLOR(expr) ...
PRINT expr ; ...

Outputs strings and numbers to the screen or other output device.

Each expr can be a constant string, a string variable or any complex expression, with commas or semicolons between each expression.

If the first expression is a device I/O channel (e.g., PRINT #1,"HELLO") the output will be sent to that device. In GRAPHICS modes other than 0 (e.g., large text GRAPHICS 2, multicolor text GRAPHICS 12, or even bitmapped graphics modes), use #6 to write to that part of the screen.

After writing the last expression, the cursor advanced to a new line, except if the statement ends in a comma, semicolon or TAB, where the cursor stays in the last position.

If there is a comma before any expression, spaces are printed to advance the printing column to the next multiple of 10, allowing easy printing of tabulated data.

The COLOR function alters the color the text that follows, until the end of the statement, depending on the graphics mode. This is abbreviated C.. Use 0 or 128 in graphics 0, for normal or inverse video. Use 0, 32, 128 or 160 in graphics mode 1 and 2 for the four available text colors, see the two examples below:

  ' In GRAPHICS 0:
  ? "NORMAL"; COLOR(128) "INVERSE"

  ' In GRAPHICS 2:
  S = 1234
  ? #6, "SCORE: "; COLOR(32) S

The TAB function advances the position to a column multiple of the argument, so that TAB(10) is the same as using a comma to separate arguments. This is abbreviated T..

The RTAB function, abbreviated RT., advances the position so that the next argument to print ends just before a column multiple of the argument, right aligning the printing of the data. This function must be immediately followed by a variable or a string to align.

Note that ,, TAB and RTAB always print at least one space, and that to separate TAB or RTAB and the previous and next arguments you can use a ; or simply a space.

See the Device Input and Output Statements section for the PRINT # usage.

This example shows the usage of TAB and RTAB, note that the columns will be left and right aligned respectively:

  FOR i=0 TO 10
    n = i*(9-2*i)*134
    ? TAB(8) "Val:" RTAB(20) n
  NEXT

The output is:

  Val:0         0
  Val:1       938
  Val:2      1340
  Val:3      1206
  Val:4       536
  Val:5      -670
  Val:6     -2412
  Val:7     -4690
  Val:8     -7504
  Val:9    -10854
  Val:10   -14740

Advanced: To implement the spacing on ,, TAB and RTAB, FastBasic uses the current column in the OS, so that POSITION and printing to a graphics screen will work ok, unlike Atari BASIC; but when printing to a file or other devices the number of spaces will not be correct. Avoid using the functions to print to any device except the screen.

Advanced: The COLOR function does an exclusive or of the given value with the value of each character in the original string before printing.

Advanced: When writing abbreviated code, you can omit the semicolon in almost all places, and just join the values together. Avoid doing this in common code for better readability.

Writes A Character To Screen
PUT num / PU.

Outputs one character to the screen, given by it's ATASCII code.

Clears The Screen
CLS

Clears the text screen. This is the same as PUT 125. For clearing the graphics screen, you can use CLS #6.

Control Statements

Endless Loops
DO
LOOP / L.

Starts and ends an endless repetition. When reaching the LOOP statement the program begins again, executing from the DO statement.

The only way to terminate the loop is via an EXIT statement.

Calls A Subroutine
**EXEC name num1, ... / EXE. / @ **

Calls the subroutine name, with the optional parameters num1 and so on, separated by commas.

Note that you must use the same number of parameters in the PROC definition, before or after the call.

Instead of EXEC you can simply use a @ in front of the procedure name. (i.e., these are equivalent: EXEC GAMEOVER and @GAMEOVER.)

Exits From Loop Or Procedure
EXIT / EX.

Exits current loop or subroutine by jumping to the end.

In case of loops, the program continues after the last statement of the loop. In case of PROC, the program returns to the calling EXEC

Loop Over Values Of A Variable
FOR var=value TO end [STEP step] / F. T. S.
NEXT var / N.

FOR loop allows performing a loop a specified number of times while keeping a counting variable.

First assigns the value to var, and starts iterations. var can be any variable name or a word array position (like "array(2)").

In each iteration, the command first compares the value of var with end, if the value is past the end it terminates the loop.

At the end of the loop, var is incremented by step (or 1 if STEP is omitted) and the loops repeats.

An EXIT statement also terminates the loop and skips to the end.

Note that if step is positive, the iteration ends when the value of var is bigger than end, but if step is negative, the iteration ends if value of var is less than end.

Also, end and step are evaluated only once at beginning of the loop; that value is stored and used for all loop iterations.

If at the start of the loop value is already past end, the loop is completely skipped.

A slightly modified usage of the FOR / NEXT loop allows for excluding the variable name from NEXT; this is required if var is an array.

This is an example of NEXT without variable:

  ' Sample of FOR/NEXT loop without
  ' NEXT variable name
  FOR i=0 to 1000 step 100
    ? i
  NEXT

Conditional Execution
IF condition THEN statement / I. T.
IF condition
ELIF condition / ELI.
ELSE / EL.
ENDIF / E.

The first form (with THEN) executes one statement if the condition is true.

This differs from Atari BASIC, TurboBASIC XL, and others, which will execute all statements after THEN until the end of the line.

For example:

  A=1
  IF A=0 THEN ? "ZERO":? "THE END"

Results in THE END being printed in FastBasic, whereas nothing would be printed in Atari BASIC.

The second form executes all statements following the IF (up until an ELIF, ELSE, or ENDIF) only if the condition is true.

If the condition is false, optional statements following the ELSE (until an ENDIF) are executed.

In case of an ELIF, the new condition is tested and acts like a nested IF until an ELSE or ENDIF.

This is an example of a multi-line IF statement:

IF _condition-1_
  ' Statements executed if
  ' _condition-1_ is true
ELIF _condition-2_
  ' Statements executed if
  ' _condition-1_ is false but
  ' _condition-2_ is true
ELIF _condition-3_
  ' Also, if _condition-1_ and
  ' _condition-2_ are false but
  ' _condition-3_ is true
ELSE
  ' Executed if all of the above
  ' conditions are false
ENDIF

Define A Subroutine.
PROC name var1 .../ PR.
ENDPROC / ENDP.

PROC statement starts the definition of a subroutine that can be called via EXEC or @.

You can pass a list of integer variables separated by spaces after the PROC name to specify a number of parameters, the variables will be set to the values passed by the EXEC call. Those variable names are always global, so the values set are seen outside the PROC.

The number of parameters in the PROC definition and in all the EXEC calls must be the same.

Note that if the PROC statement is encountered while executing surrounding code, the full subroutine is skipped, so PROC / ENDPROC can appear any place in the program.

Loop Until Condition Is True
REPEAT / R.
UNTIL condition / U.

The REPEAT loop allows looping with a condition evaluated at the end of each iteration.

Executes statements between REPEAT and UNTIL once, then evaluates the condition. If false, the loop is executed again, if true the loop ends.

An EXIT statement also terminates the loop and skips to the end.

Loop while condition is true
WHILE condition / W.
WEND / WE.

The WHILE loop allows looping with a condition evaluated at the beginning of each iteration.

Firstly it evaluates the condition. If false, it skips the whole loop to the end. If true, it executes the statements between WHILE and WEND and returns to the top to test the condition again.

An EXIT statement also terminates the loop and skips to the end.

Graphic and Sound Statements

Set Color Number
COLOR num / C.

Changes the color of PLOT, DRAWTO and the line color on FILLTO to num.

Draws A Line
DRAWTO x, y / DR.

Draws a line from the last position to the given x and y positions.

Sets Fill Color Number
FCOLOR num & FC.

Changes the filling color of FILLTO operation to num.

Fill From Line To The Right
FILLTO x, y / FI.

Draws a line from the last position to the given x and y position using COLOR number. For each plotted point it also paints all points to the right with the FCOLOR number, until a point with different color than the first is reached.

Sets Graphic Mode
GRAPHICS num / G.

Sets the graphics mode for graphics operations. Below is a basic chart of GRAPHICS modes, their full screen resolution and number of available colors.1

Text modes234:

Mode Resolution # Of Colors
GR. 0 40x24 2
GR. 1 20x24 5
GR. 2 20x12 5
GR. 12 40x24 5
GR. 13 40x12 5

Bitmapped graphics modes:5

Mode Resolution # Of Colors
GR. 3 40x24 4
GR. 4 80x48 2
GR. 5 80x48 4
GR. 6 160x96 2
GR. 7 160x96 4
GR. 8 320x192 2
GR. 9 80x192 16 shades
GR. 10 80x192 9
GR. 11 80x192 16 hues
GR. 14 160x192 2
GR. 15 160x192 4

For graphics modes which include a 4-line GRAPHICS 0 style text window at the bottom (all but 0, 9, 10, and 11), add 16 to the mode number to disable the text window. (e.g., GRAPHICS 2+16)

Add 32 to the mode number to prevent the graphics data from being cleared. (Note: Some graphics data may be replaced when changing modes.)

Advanced: The Atari OS S: screen device dictate which GRAPHICS modes are available. However (as demonstrated by the text window), the Atari can mix graphics modes via use of Display Lists. The ANTIC graphic chip uses a different set of values to reflect the different graphics modes (and modes 9, 10 and 11 utilize a feature managed by the GTIA chip), as well as other features (blank scan-lines, fine scrolling, Display List Interrupts, etc.) Consult De Re Atari chapter 2, "ANTIC and the Display List" for more details.

Get color of pixel
LOCATE x, y, var / LOC.

Reads the color of pixel in the specified x and y coordinates and store into variable var.

Plots A Single Point
PLOT x, y / PL.

Plots a point in the specified x and y coordinates, with the current COLOR number.

Player/Missile Graphics Mode
PMGRAPHICS num / PMG.

Set up Atari Player / Missile graphics. A value of 0 disables all player and missiles; a value of 1 sets up single line resolution; a value of 2 sets up double line resolution.

Single line mode uses 256 bytes per player, while double line uses 128 bytes per player. (Note that all four missiles share the same data.)

For retrieving the memory address of the player or missile data use the PMADR() function.

Player/Missile Horizontal Move
PMHPOS num,pos / PM.

Set the horizontal position register for the player or missile num to pos.

Players 0 to 3 correspond to values 0 to 3 of num; missiles 0 to 3 correspond to the values 4 to 7, respectively.

This is the same as writing: POKE $D000 + num , pos

Note: Player/Missile graphics on the Atari are strips that are as tall as the screen, and therefore to move a shape vertically its data must be moved within their 128- or 256-byte buffer (using the MOVE statement, for example).

Sets Displayed Color
SETCOLOR num, hue, lum / SE.

Alters the color registers so that color number num has the given hue and luminance.

To set Player/Missile colors use negative values of num, -4 for player 0, -3 for player 1, -2 for player 2, and -1 for player 3.

Missiles share the same color as their player, unless you combine them into a "5th Player" by setting bit number 4 of the GPRIOR register, e.g.: POKE 623,16. (You must also move them horizontally in unison if you wish to use them as a true 5th Player.)

It is possible to cause pixels of certain overlapping players to produce a third color (or black) by setting bit number 5 of the GPRIOR register, e.g. POKE 623,32.

Consult the GPRIOR section of Mapping the Atari for more details.

Adjust Voice Sound Parameters
SOUND voice, pitch, dist, vol / S.
SOUND voice
SOUND

Adjusts sound parameters for voice (from 0 to 3) of the given pitch, distortion and volume.

If only the voice parameter is present, that voice is cleared so no sound is produced by that voice.

If no parameters are given, it clears all voices so that no sounds are produced.

Note: TurboBASIC XL offers a DSOUND statement to pair sound channels for increased (16-bit) frequency range. This is not available in FastBasic.

Device Input and Output Statements

Binary read from file
BGET #iochn,address,len / BG.

Reads length bytes from the channel iochn and writes the bytes to address.

For example, to read to a byte array, use ADR(array) to specify the address.

On any error, ERR() will hold an error code, on success ERR() reads 1.

Binary Read From File
BPUT #iochn,address,len / BP.

Similar to BPUT, but writes length bytes from memory at address to the channel iochn.

On any error, ERR() will hold an error code, on success ERR() reads 1.

Close Channel
CLOSE #iochn / CL.

Closes the input output channel iochn, finalizing all read/write operations.

On any error, ERR() will hold an error code, on success ERR() reads 1.

Note that it is important to read the value of ERR() after close to ensure that written data is really on disk.

Reads bytes from file
GET #iochn, var, ...

Reads one byte from channel iochn and writes the value to var.

var can be a variable name or an array position (like array(123))

In case of any error, ERR() returns the error value.

Input Variable Or String From File
INPUT #iochn, var / IN.

Reads a line from channel iochn and stores to var.

If var is a string variable, the full line is stored.

If var is a numeric variable, the line is converted to a number first.

On any error, ERR() will hold an error code, on success ERR() reads 1.

Opens I/O Channel
OPEN #ioc,mode,ax,dev / O.

Opens I/O channel ioc with mode, aux, over device dev.

To open a disk file for writing, mode should be 8, aux 0 and dev the file name as "D:name.ext".

To open a disk file for reading, mode should be 4, aux 0 and dev the file name as "D:name.ext".

See Atari BASIC manual for more documentation in the open modes, aux values, and device names.

On any error, ERR() will hold an error code, on success ERR() reads 1.

Print Strings And Numbers To A File
PRINT #iochn, ... / ?

Uses the same rules as the normal print, but all the output is to the channel iochn. Note that you must put a comma after the channel number, not a semicolon.

On any error, ERR() will hold an error code, on success ERR() reads 1.

Note that you can only read the error for the last element printed.

Outputs One Byte To The File
PUT #iochn, num / PU.

Outputs one byte num to the channel iochn.

On any error, ERR() will hold an error code, on success ERR() reads 1.

Generic I/O Operation
XIO #iochn, cmd, aux1, aux2, dev / X.

Performs a general input/output operation on device dev, over channel ioc, with the command cmd and auxiliary bytes aux1 and aux2.

Note that the arguments of XIO statements are in different order than Atari BASIC, for consistency with other statements the iochn is the first argument.

Example: to delete the file "FILE.TXT" from disk, you can do:

  XIO #1, 33, 0, 0, "D:FILE.TXT"

General Statements

Line comments
' / .

Any line starting with a dot or an apostrophe will be ignored. This is analogous to REM in Atari BASIC.

Clears variables and free memory
CLR

Clears all integer and floating-point variables to 0, all strings to empty strings and frees all memory associated with arrays.

After CLR you can't access arrays without allocating again with DIM.

Defines array with initial values
DATA arr() [type] = n1,n2, / DA.

This statement defines an array of fixed length with the values given.

The array name should not be used before, and type can be BYTE (abbreviated B.) or WORD (abbreviated W.). If no type is given, a word data is assumed.

If you end the DATA statement with a comma, the following line must be another DATA statement without the array name, and so on until the last line.

Example:

  DATA big() byte = $12,$23,$45,
  DATA       byte = $08,$09,$15

Note that the array can be modified afterwards like a normal array.

Advanced Usage

Byte DATA arrays can be used to include assembler routines (to call via USR, see the example above), display lists and any other type of binary data.

To facilitate this, you can include constant strings and the address of other byte DATA array by name.

All the bytes of the string, including the initial length byte are included into the DATA array.

Example:

  DATA str() B. = "Hello", "World"
  X = ADR(str)
  ? $(X), $(X+6)
  DATA ad() B. = $AD,str,$A2,0,$60
  ? USR(ADR(ad)), str(0)

Loading data from a file

The cross-compiler also supports loading data from a file directly into the program, using the BYTEFILE (abbreviated BYTEF.) and WORDFILE (abbreviated WORDF. or simply F.) types and a file name enclosed in double quotes.

Example:

  DATA img() bytefile "img.raw"
  DATA pos() wordfile "pos.bin"

The compiler will search the file in the same folder than the current basic source.

Storing data into ROM

In addition to the above, the cross compiler allows to specify that the data should be stored in ROM, instead of the default in RAM. This means that the data can't be modified in targets that use ROM (cartridges), but will lower RAM usage.

To specify this, simply add the ROM word after the type:

  DATA img() ROM 1234,5678
  DATA pos() BYTE ROM 1,2,3,4

Decrements variable by 1
DEC var / DE.

Decrements the variable by 1; this is equivalent to "var = var - 1", but faster.

var can be any integer variable or integer array element.

Allocate an Array / Define Var
DIM arr(size) [type], .../ DI.
DIM var, var$, var% ...

The DIM statement allows defining arrays of specified length, and declaring variables explicitly, without assigning a value.

The type must be BYTE (abbreviated B.) to define a byte array, with numbers from 0 to 255, or WORD (can be left out) to define an array with integers from -32768 to 32767.

If the name arr ends with a $ or a % symbol, this defines a string array or floating point array respectively, in this case you can't specify a type.

The size of the array is the number of elements plus one, the elements are numerated from 0, so that an array dimensioned to 10 holds 11 values, from 0 to 10.

The array is cleared after the DIM, so all elements are 0 or an empty string.

In the second form, the variables given in the list are defined with the correct type, without giving a default value. The variables can be defined multiple times without an error if the types are always the same.

You can DIM more than one array or variable by separating the names with commas.

Example:

  DIM A(10), X, T$
  ? A(5), X

Ends Program
END : Ends program.

Terminates current program. END is only valid at end of input.

Increments Variable By 1
INC var

Increments the variable by 1, this is equivalent to "var = var + 1", but faster.

var can be any integer variable or integer array element.

Pauses Execution
PAUSE num / PA. PAUSE

Stops the current execution for the specified amount of time.

num is the time to pause in "jiffies", this is the number of TV scans in the system; 60 per second in NTSC or 50 per second in PAL.

Omitting num is the same as giving a value of 0, and pauses until the vertical retrace. This is useful for synchronization to the TV refresh and for fluid animation.

Resets internal timer
TIMER/ T.

Resets value returned by TIME function to 0.

Floating Point Statements

Those statements are only available in the floating point version.

Sets "degrees" mode
DEG

Makes all trigonometric functions operate in degrees, so that 360 is the full circle.

Sets "radians" mode
RAD

Makes all trigonometric functions operate in radians, so that 2pi is the full circle.

This mode is the default on startup.

Low Level Statements

These are statements that directly modify memory. Use with care!

Writes a 16bit number to memory
DPOKE address, value / D.

Writes the value to the memory location at address and address+1, using standard CPU order (low byte first).

Copies Bytes In Memory
MOVE from, to, length / M.
-MOVE from, to, length / -.

Copies length bytes in memory at address from to address to.

The MOVE version copies from the lower address to the upper address; the -MOVE version copies from upper address to lower address.

The difference between the two MOVE statements is in case the memory ranges overlap; if from is lower in memory than to, you need to use -MOVE, else you need to use MOVE, otherwise the result will not be a copy.

MOVE a, b, c is equivalent to:

  FOR I=0 to c-1
    POKE b+I, PEEK(a+I)
  NEXT I

On the other hand, -MOVE a, b, c is instead:

  FOR I=c-1 to 0 STEP -1
    POKE b+I, PEEK(a+I)
  NEXT I

Sets Memory To A Value
MSET address, length, value / MS.

Writes length bytes in memory at given address with value.

This is useful to clear graphics or P/M data, or simply to set an string to a repeated value.

MSET a, b, c is equivalent to:

  FOR I=0 to b-1
    POKE a+I, c
  NEXT I

Writes A Byte To Memory
POKE address, value / P.

Writes the value (modulo 256) to the memory location at address.

Display List Interrupts

Note: This is an advanced topic.

Display list interrupts (normally called DLI) are a way to modify display registers at certain vertical positions on the screen.

You can use them to:

  • Display more colors in the image, by changing color registers - registers from $D012 to $D01A.

  • Split one Player/Missile graphics to different horizontal positions - registers from $D000 to D007.

  • Change scrolling position, screen width, P/M width, etc.

FastBasic allows you to specify one or more DLI routines, activate one or deactivate all DLI by using the DLI statement:

Define a new DLI
DLI SET name = op1, op2, ... / DLIS.

Setups a new DLI with the given name and performing the op operations.

Each operation is of the form: data INTO address or data WSYNC INTO address.

data is one constant byte or the name of a DATA BYTE array, and address is a memory location to modify.

If data is a DATA array, the first element (at index 0) will be used at the first line with DLI active in the screen, the second element at the second active line, etc.

The WSYNC word advances one line in the display area (this is done by writing to the WSYNC ANTIC register), so the value is set in the next screen line. You can put the WSYNC word multiple times to advance more than one line. This allows one DLI to modify multiple lines at the screen.

Multiple INTO words can be used to write more than one register with the same value.

INTO can be abbreviated to I. and WSYNC to W..

You can specify any number of operations, but as each one takes some time you could see display artifacts if you use too many.

Note that by defining a DLI you are simply giving it a name, you need to activate the DLI afterwards.

You can split a DLI definition over multiple lines, just like DATA by ending a line with a comma and starting the next line with DLI =

Enable a DLI
DLI name / DL.

This statement enables the DLI with the given name, the DLI must be defined before in the program.

This setups the OS DLI pointer to the named DLI and activates the interrupt bit in the display processor (the ANTIC chip), but does not activates on which lines the DLI must be called.

To define on which lines the DLI is active you must modify the Display List, see the example at the end of the section.

You can also pass the name of a DATA BYTE array with a custom machine language routine to the DLI statement, the routine must begin with a PHA and end with PLA and RTI.

Disable a DLI
DLI / DL.

This statement simply disables the DLI, returning the display to the original

DLI Examples

This is the most basic example of a DLI that simply changes the background color at the middle of the screen:

  ' Define the DLI: set background
  ' color to $24 = dark red.
  DLI SET d1 = $24 INTO $D01A
  ' Setups screen
  GRAPHICS 0
  ' Alter the Display List, adds
  ' a DLI at line 11 on the screen
  POKE DPEEK(560) + 16, 130
  ' Activate DLI
  DLI d1
  ' Wait for any keyu
  ? "Press a Key" : GET K
  ' Disable the DLI
  DLI

The next example shows how you can use a DLI to change multiple values in the screen:

  ' An array with color values
  DATA Colors() BYTE = $24,$46,$68
  ' Define the DLI: set background
  ' color from the Color() array
  ' and text back color with value
  ' $8A in the same line and then
  ' the black in to the next line.
  DLI SET d2 = Colors INTO $D01A,
  DLI        = $8A INTO $D018,
  DLI        = $00 WSYNC INTO $D018
  ' Setups screen
  GRAPHICS 0
  ' Adds DLI at three lines:
  POKE DPEEK(560) + 13, 130
  POKE DPEEK(560) + 16, 130
  POKE DPEEK(560) + 19, 130
  ' Activate DLI
  DLI d2
  ' Wait for any keyu
  ? "Press a Key" : GET K
  ' Disable the DLI
  DLI

The final example shows how you can move multiple P/M using one DLI

  ' Player shapes, positions and colors
  DATA p1() BYTE = $E7,$81,$81,$E7
  DATA p2() BYTE = $18,$3C,$3C,$18
  DATA pos() BYTE = $40,$60,$80,$A0
  DATA c1() BYTE = $28,$88,$C8,$08
  DATA c2() BYTE = $2E,$80,$CE,$06
  ' Our DLI writes the position and
  ' colors to Player 1 and Player 2
  DLI SET d3 = pos INTO $D000 INTO $D001,
  DLI        = c1 INTO $D012, c2 INTO $D013
  GRAPHICS 0 : PMGRAPHICS 2
  ' Setup our 4 DLI and Players
  FOR I = 8 TO 20 STEP 4
    POKE DPEEK(560) + I, 130
    MOVE ADR(p1), PMADR(0)+I*4+5,4
    MOVE ADR(p2), PMADR(1)+I*4+5,4
  NEXT
  ' Activate DLI
  DLI d3
  ? "Press a Key"
  REPEAT
    PAUSE
    pos(0) = pos(0) + 2
    pos(1) = pos(1) + 1
    pos(2) = pos(2) - 1
    pos(3) = pos(3) - 2
  UNTIL KEY()
  DLI

Some useful registers

This is a table of some useful registers to change during a DLI:

Address Register
$D000 Player 0 horizontal pos.
$D001 Player 1 horizontal pos.
$D002 Player 2 horizontal pos.
$D003 Player 3 horizontal pos.
$D004 Missile 0 horizontal pos.
$D005 Missile 1 horizontal pos.
$D006 Missile 2 horizontal pos.
$D007 Missile 3 horizontal pos.
$D012 Color of player/missile 0
$D013 Color of player/missile 1
$D014 Color of player/missile 2
$D015 Color of player/missile 3
$D016 Color register 0
$D017 Color register 1
$D018 Color register 2
$D019 Color register 3
$D01A Color of background

Atari SIO Statements

The Atari Serial Input Output interface is the low-level interface between the Atari 8-bit computers and the serial peripherals, like disk-drives and modems.

Send any command over SIO
SIO ddevic, dunit, dcomnd, dstats, dbuf, dtimlo, dbyt, daux1, daux2

This function can be used to send any SIO command to any SIO device. For example, this command is used to read or write one sector in a floppy disk, or send special commands to a FujiNet network device.

Parameter Description
DDEVIC Device # (e.g. $71)
DUNIT Unit #
DCOMND Command # ($00-$FF)
DSTATS Read($40) / Write($80)
DBUF Target buffer address
DTIMLO Timeout value
DBYT # of bytes in payload
DAUX1 First Aux parameter
DAUX2 Second Aux parameter

The meanings of each of these is highly dependent on the target device.

Get last SIO error function
SERR() / SE.

This function returns the value in DSTATS, which contains the error of the last SIO operation from the device.

In the context of the FujiNet device, can be used, along with DVSTAT+4 to determine any error from a network operation.

FujiNet Statements

NOTE: The FujiNet Statements are not available in the integer-only version.

These are statements that talk to the FujiNet network adapter, and can be used to open network connections, using any protocol supported.

Each of these statements require a unit number, of which 8 are available, numbered 1-8.

The general flow of use is:

  • NOPEN a connection
  • In a loop
    • Check for any traffic with NSTATUS
    • NGET if needed
    • Send any traffic with NPUT
  • When done, NCLOSE.

Open a Network Connection
NOPEN unit, mode, trans, url / NO.

Uses N: unit to open a connection to url using the desired mode and trans settings.

Example URLs might be: N:HTTPS://www.gnu.org/licenses/gpl-3.0.txt

Common modes:

  • 4: READ, mapped e.g. to GET in HTTP
  • 6: DIRECTORY, e.g. PROPFIND in HTTP
  • 8: WRITE, mapped e.g. to PUT in HTTP
  • 12: READ/WRITE, e.g. for TCP
  • 13: Mapped to POST in HTTP

Common trans:

  • 0: No translation of characters.
  • 1: Change CR to ATASCII EOL.
  • 2: Change LF to ATASCII EOL.
  • 3: Change CR and LF to EOL.

Close a Network Connection
NCLOSE unit / NC.

Closes a network connection unit previously opened by NOPEN.

Get Network Connection Status
NSTATUS unit / NS.

Queries the status of specified network unit. The result is stored in DVSTAT starting at $02EA, and has the format:

Address Description
$02EA # of bytes waiting (LO)
$02EB # of bytes waiting (HI)
$02EC Connected? (0 or 1)
$02ED Most recent error #

You can easily get the # of bytes waiting by doing the following:

NSTATUS 1
BW = DPEEK($02EA)

Read Bytes from Network to addr
NGET unit, addr, len / NG.

Write bytes to Network from addr
NPUT unit, addr, len / NP.

These two functions are complements of each other, reading and writing len bytes to and from addr as needed.

When reading, len must be less than, or equal to the number of bytes waiting to be received, or an SIO error will result. Therefore, it is a good idea to figure out how many bytes are waiting using the NSTATUS command.

Conversely, when writing, len must be less than, or equal to the number of bytes in the source buffer.

For example all of the available SIO commands for FujiNet Network at this link: SIO Commands for FujiNet Devices

Footnotes

  1. GRAPHICS 0 and GRAPHICS 8 offer two colors, where the "on" pixels may be a different shade (luminance) of the background color's hue, but cannot have its own hue. (Television color artifacting effects can be utilized to simulate two additional colors.) Use SETCOLOR 2,H,L1 and SETCOLOR 1,0,L2 (or POKE 710,H*16+L1 & POKE 709,L2). ↩

  2. Mode 0 (and the text window found at the bottom of most other modes) can render 128 different characters (from a character set, aka font) in both normal video, and inverse video, based on whether the high bit of the character is set. See PRINT COLOR(). ↩

  3. Modes 1 and 2 are text modes that offer multiple colors, but only a single color (plus the background) may be used by any given character cell. The colors are chosen by the two high bits of the character. This means only half of a character set (font) -- 64 shapes -- may normally be used. See PRINT COLOR(). ↩

  4. Modes 12 and 13 are multicolor text modes, where every pair of two bits in a character's bitmap data are used to represent one of four colors. As with mode 0, 128 characters may be used. However, when the high bit is set (which produces an inverse-video effect in mode 0), the effect in these modes is to change which color palette register is used for the fourth color (pixels comprised of 11 bits); instead of SETCOLOR 2 (aka POKE 710), the color from SETCOLOR 3 (aka POKE 711) will be used. See PRINT COLOR(). ↩

  5. The so-called "GTIA modes" -- 9, 10, and 11 -- offer 16 shades of the given background color (use SETCOLOR 4,H,0 or POKE 712,H*16), all nine color registers (SECTOLOR N,H,L or POKE 704+N,H*16+L), or 15 hues of a particular brightness (the background remains darkest; use SETCOLOR 4,0,L), respectively. ↩