Word Writer 6 Commodore 64 Source Code | Hacker News Reader

Word Writer 6 Commodore 64 Source Code

github.com

112 pointsindigodaddy20 comments

I'd never heard of Word Writer before. I found the manual on archive.org [1]. I love how the algorithm for how the spell checker works is described by the manual! Also that's one fancy manual for the 1980s - they even have icons that represent the keystrokes in the manual.

[1] https://archive.org/details/Word_Writer_6_Users_Manual_1991_...

An Apple manual from the end of the 70s:

https://archive.org/details/The_Applesoft_Tutorial_HQ_color/...

And they only got fancier from there. If anything, current manuals and books are a lot less frilly than your typical early personal computer manual.

The manual is also in the GitHub repo: https://github.com/jefflomax/word-writer-6-commodore-64/blob...

> Also that's one fancy manual for the 1980s

Note that this version of the manual was published in 1991.

Getting a spell checker working on a machine with 64kB of RAM and a very slow disk with only 170kB of storage is black magic. Start thinking about how you would do it and check the size of the dictionary and it seems like a miracle to get it working at all.

I wrote the spell checker for WordPro 64. I had to invent a data compression algorithm based on something that I read in an Encyclopedia Brown story to fit the 70k word dictionary onto that 170k floppy. I also wrote one of the original Fastload cartridges (the GT4, also sold by Pro Line software) which was bundled with WordPro 64. So I’m pretty familiar with this problem domain :)

I wrote Word Writer for Commodore 64. Creating the spelling checker was quite the task. I got a PC with enough RAM for a large RamDisk, managed to write an analysis tool that quantified all 2, 3, and 4 letter groups in the dictionary. This was encoded into 5 bit words, (no uppercase), serialized, and as it streamed alphabetically thru the dictionary, each entry kept the leading letters from the prior one. It worked, after a lot of debugging.

I've made the source code to Word Writer open source. Please encourage your programmer friends to open source off copyright works. The software industry is very eager to claim copyright protection, but generally fails to deliver that literary work to the public as copyright is intended.

Also - a lot of old software is trapped in tooling that isn't available. In my case that was the 2500AD cross assembler. So I wrote another tool (also open source https://github.com/jefflomax/converttoacme) that converts the software from the 2500AD format to the wonderful, modern, open ACME cross assembler. That's the reason WHY you can download WW6, open it in VSCode, add a few free extensions, assemble it, and shoot it right into the fantastic VICE Commodore Emulator and see it run.

OH, and if you think the spelling checker was difficult - think about the Thesaurus and 55K words packed onto a 174MB disc! That thing had a host of different word sizes being streamed thru.

Hope you enjoy this, feel free to ping.

Jeff Lomax

P.S. I have a source listing of Partner 64, but it's on fan fold paper, no media. I'd love to know a good way to OCR it without destroying the fanfold listing...

> P.S. I have a source listing of Partner 64, but it's on fan fold paper, no media. I'd love to know a good way to OCR it without destroying the fanfold listing...

How many pages is it? You can probably take pictures using a camera, or even a mobile phone.

Good job releasing this source code.

This will ensure it's preserved properly. It might even spawn new work fixing bugs or adding features.

Much better than letting the media with the sources rot somewhere, eventually lost.

Wrote a lot of term papers in Word Writer (printed on a Star NX-10C). This is a delightful thing to see.

Word Writer 2 and a Commodore MPS-801 printer got me through high school.

Word Writer 4 and an Okidata Okimate 20 printer, here. Some years ago I wrote a quick-and-dirty program to convert the files to HTML [1]... only to be absolutely horrified at how awful my writing was as a teenager. ;)

[1] https://cafbit.com/post/migrating_legacy_documents_from_word...

The MPS-801 was a uni-directional printer, with a coil spring to pull the head back to the left hand side. Other than making it slow, it also flew back via the spring and hit the left side with a loud thunk after ever line, which made printing stuff out late at night when everyone else was sleeping, very problematic.

In high school, if needed, I was always able to turn on the charm and talk my way into being able to turn papers in late without penalty, so I was never up late enough printing a paper to disturb anyone. I left my C64 behind when going to college so the MPS-801 never disturbed anyone there either.

I've never seen a word processor other than Compute!'s Speedscript that implemented Commodore-Z, which transposed the last two characters typed.

Very useful for people who often sometimes type words like "its" as "ist."

Good times. I used Speedscript as a pro tech writer for a few years, connected to a Brother typewriter with a parallel port interface dongle thing. Typed it in myself.

    ; THIS IS REMOVED BECAUSE IT KILLS FILES OPENED FOR
    ; MAIL-MERGE.  I DON'T KNOW RIGHT NOW WHY THIS IS HERE

Ah, a classic from before the days of source control.

I see stuff like this in source controlled repos all the time! The argument is that, even if it's in the git history, it'll get lost/forgotten about in case it's ever needed again.

In some(!) cases it totally makes sense to provide a hint for the person who might read the source code and wonder why there is a certain 'void' recognisable in the code, because it has been designed around another piece of code or data that is no longer there. Better to keep a short comment in the source file for readers to stumble over, and then via blame it's easy to read up on the whole history in version control.

(2019)- pretty cool though - I hadn't heard about this release till now.

Reply on news.ycombinator.com