This repository has the source code for the SpeedScript word processor by Charles Brannon for the Commodore 64.
The "original" source as shown in the book in is the original directory. Also in the original directory are monolithic versions converted to ca65 format. The code in src has been split up into different modules, but by using segments it generates the exact same matching binary as originally released for SpeedScript 3.2r2. It is a bit clumsy right now, but is my first attempt and should get cleaned up a bit over time.
More work is needed to add patches and use conditional assembly. This will allow us to build the original binaries as well as versions with certain patches. For example, building SpeedScript-80 should be fairly straight-forward, while building SpeedScript 128 would be more involved. Short patches are more readily reverse engineered and integrated. I'm not aware of any source for SpeedScript-80, SpeedScript 128, or SpeedScript 128 Plus so it will have to be reverse engineered from a disassembly.
Note that SpeedScript-80 is a set of patches for 80 column mode on a C128 in C64 mode. There is a Instant 80 patch for the C64 that provides a software based 80 column preview with the print to screen feature.
The SpeedScript word processor was very popular as it was made freely available as a "type-in" program in COMPUTE! Magazine and COMPUTE! Gazette. It was available for the Commodore 64/128, VIC-20, Atari 800, and Apple II.
This repository has the source code as published by COMPUTE! Magazine in a book made up of several articles from the magazines. The source code was not made available in the magazines. Primarily due to its length I would assume and the fact that most end users wouldn't be able to use it. They were interested in using the word processor directly and wouldn't have the PAL Assembler.
The book is available here: https://archive.org/details/Computes_Speedscript
I could not find this source code anywhere on the Internet (other than the scans at archive.org) so I went ahead and typed it all in as close to the original printed form as possible. Look in the 'original' directory.
After proofreading it several times (and regularly catching my typing errors) I converted it to CA65 format. CA65 is the assembler portion of the CC65 cross compiler. This is the same tool used for the reverse engineered GEOS source code. I hope to improve upon my work here and modularize the source code and add the ability to conditionally assemble additional features. Many additional features were released as patches and I would like to incorporate them here as assembler source code.
You can build SpeedScript by running make:
$ makeYou can also just run the tool directly on monolithic sources:
$ cl65 --cpu 6502x -o speedscript32r2.prg -t c64 -C c64-asm.cfg original/speedscript32r2.sRunning make also builds speedscript-i80.prg which integrates the Instant 80 patch that provides a software 80 column (or soft-80) preview to screen. It also builds speedscript-ec.prg which integrates the Easy Cursor patch. The speedscript-i80-ec.prg file combines both Instant 80 and Easy Cursor in one program.
Please submit a pull request, open an issue, or otherwise contact me for improvements to the build process.
Transfer the speedscript.prg or speedscript-80.prg file to your Commodore 64 or emulator and load it.
READY.
LOAD"SPEEDSCRIPT",8
SEARCHING FOR SPEEDSCRIPT
LOADING
READY.
RUN
Here is a quick reference table of SpeedScript keystrokes.
| Keystroke | Function |
|---|---|
| CTRL A | Change case (of letter under cursor) |
| CTRL B | Change border color |
| CTRL D | Delete (Sentence, Word, Paragraph) (to the left) |
| CTRL E | Erase (Sentence, Word, Paragraph) (to the right) |
| CTRL G | Global search and replace |
| CTRL H | Hunt for phrase with SHIFT: Select hunt phrase |
| CTRL I | Enter/exit insert mode |
| CTRL J | Replace with SHIFT: Select replace phrase |
| CTRL K | Kill buffer |
| CTRL L | Change text character color |
| CTRL P | |
| CTRL R | Restore buffer |
| CTRL V | Verify |
| CTRL X | Transpose characters |
| CTRL Z | Go to end of text |
| CTRL = | Display amount of free memory |
| CTRL ⇧ | Send disk command or read error channel |
| CTRL 4 | Display disk directory |
| CTRL £ | Enter format (printer) commands |
| CTRL 3 | Commodore 64 only: Same as CTRL-£ |
| SHIFT CTRL P | Print to (Screen, Disk, Printer) with options |
| CLR/HOME | Press once to go to top of screen Hold down to go to top of text with SHIFT: Erase all text |
| CRSR (left/right) | Move the cursor left one character with SHIFT: Move the cursor right one character |
| CRSR (up/down) | Go to next sentence with SHIFT: Go to previous sentence |
| RUN/STOP | Indent 5 spaces with SHIFT: Insert 255 spaces |
| RESTORE | Exit SpeedScript (Commodore 64) with RUN/STOP: Exit SpeedScript (VIC-20) |
| ⇦ | Backspace with CTRL: Delete character under cursor and close up text with SHIFT and CTRL: Delete all spaces from cursor to next character |
| RETURN | Return mark with SHIFT: End paragraph, add an extra return mark, and indent next paragraph |
| INST/DEL | Delete character with SHIFT: Insert space |
| F1 | Next Word |
| F2 | Previous Word |
| F3 | Next Sentence |
| F4 | Previous Sentence |
| F5 | Next Paragraph |
| F6 | Previous Paragraph |
| F7 | Load |
| F8 | Save |
I would like to add other versions as well as various patches. I'm not that familiar with the patches, but now that I have the base code working I can investigate. The 80 column preview mode and cursor handling seem like fairly desirable patches. Instant 80 and Easy Cursor have been integrated. I'm still looking for other official (published in magazines) and unofficial patches.
Potential patches:
- 80 column mode for C128
80 column preview for C64 (screen "printing" in 80 columns)Cursor improvements (Easy Cursor patch for 3.2)- SpeedScript 128 (major rewrite)
- SpeedScript 128 Plus
- Your favorite patch...
Please submit a pull request, open an issue, or otherwise contact me if you have source code for a patch.
The information below is an excerpt from the original printed book.
The following articles were originally published in COMPUTE! magazine, copyright 1985, COMPUTE! Publications, Inc.: "SpeedScript 3.0: All Machine Language Word Processor For Commodore 64" (March), "SpeedScript 3.0: All Machine Language Word Processor For Expanded VIC-20" (April), and "ScriptSave" (May). Copyright 1985, COMPUTE! Publications, Inc. All rights reserved Reproduction or translation of any part of this work beyond that permitted by Sec- tions 107 and 108 of the United States Copyright Act without the permission of the copyright owner is unlawful. Printed in the United States of America 10 987654321 ISBN 0-942386-94-9 COMPUTE! Publications, Inc., Post Office Box 5406, Greensboro, NC 27403, (919) 275-9809, is one of the ABC Publishing Companies and is not associated with any manufacturer of personal computers. Commodore 64 and VIC-20 are trademarks of Commodore Electronics Limited.
Below is the type-in introduction offered in the book by the original author. He explains important details about the design of the editor and the layout of the source code.
Commodore 64
Source Code
The source code for SpeedScript was originally developed on the PAL assembler (Pro-Line) and -- except for the .ASC and .WORD pseudo-op -- is compatible with the LADS assembler from The Second Book of Machine Language (COMPUTE! Books, 1984). Line numbers are omitted. Most pseudo-ops are in stan- dard MOS 6502 notation: *= updates the program counter (some assemblers use .ORG, .DB, or .DW instead); .BYT or .BYTE assembles a list of numbers; .WOR or .WORD assem- bles a list of addresses into low byte/high byte format; .ASC is used to assemble an ASCII character string (many assem- blers -- including LADS -- use .BYTE for this also); < extracts the low byte of a 16-bit expression; > extracts the high byte of a 16-bit expression (some assemblers reverse the use of < and >; others use &255 and /256 to achieve the same effect); and = is used to assign an express to a label (some assemblers use .EQU).
Beginners should make sure the undersand indirect ,y addressing, as in LDA ($FB),Y or LDA (CURR),Y. This mode is used extensively in SpeedScript.
Notice that a small portion of SpeedScript is listed in lowercase. This is how it would actually appear on your screen. It doesn't really matter which mode you're in when typing in the rest of SpeedScript -- just don't SHIFT to get uppercase.
The VIC version of SpeedScript was translated from the 64 source code and developed on the 64. There isn't room to in- clude it here, but it is very similar. Address $BDCD on the 64 becomes $DDCD on the VIC. References to location 1 (which maps in and out ROM in the 64) would be omitted for the VIC. The REFRESH routine, TOPCLR, and a few other rou- tines were changed. The WINDCOLR variable was changed to a subroutine, and the HIGHLIGHT and DELITE routines (which turn on or off the raster interrupt that highlights the command line) were removed. But about 95 percent of the source code did not need to be changed. In fact, the transla- tion only took a single day to get running, and about a week to test and debug.
SpeedScript is written in small modules. Some people think that subroutines are useful only when a routine is called more than once. I strongly believe in breaking up a problem into a number of discrete tasks. These tasks can be written as subroutines, then tested individually. Once all the modules are working, just link them together with JSRs and you have a working program.
I've also tried to use meaningful labels, but sometimes one just runs out of imagination. Comments are added below as signposts to guide you through the source code (you needn't type them in -- if you do, precede each comment with a semicolon for the sake of your assembler). Modules are also set apart with blank lines. Notice that some modules are used in rather creative ways. For example, word left/word right is used both for moving the cursor and in delimiting a word to be erased in the erase mode. Also, note that memory locations are sometimes used instead of meaningful labels. In order to fit the complete source code in memory at once, I sometimes had to compromise readability for the sake of brevity.
Crucial to the understanding of SpeedScript is the RE- FRESH routine. Study it carefully. REFRESH is the only time SpeedScript writes directly to the screen (Kernal ROM routine $FFD2 is used to print on the command line). It automatically takes care of word-wrap and carriage returns, and provides useful pointers so that the CHECK routine can easily scroll the screen. This frees the rest of SpeedScript to just move and modify contiguous memory. Carriage returns are not padded out in memory with spaces to fill the rest of a line; the RE- FRESH routine takes care of this transparently.
Also, for the sake of compact code, Kernal and BASIC routines are used heavily for routines like Save and Load and for printing numbers.
You'll see some references to location 1, used for mapping the ROMs in and out of the address space. SpeedScript stores the main text from the end of the program all the way up to the beginning of I/O space ($CF00). One page of memory is used as a boundary between text areas (the text buffer starts at $D000). This may seem superstitious, but it provides for a margin of error. BASIC is mapped back in when SpeedScript needs to call $BDCD to print a number, and then mapped back out. The Kernal ROM is left mapped in, since it is con- stantly called, but it's mapped out when the program needs to write to or read from the buffer, which is stored beneath the I/O area and the Kernal. Refer to the memory map shown on page 126.