savetz / appleiibot

The code that runs the Apple II Twitter bot at https://twitter.com/AppleIIBot

AppleIIBot

The code that runs the Apple II Twitter bot at https://twitter.com/AppleIIBot

I'm sharing this so people can use this as a stepping stone to making their own, different bots.

Documentation for using the bot is at https://appleiibot.com

The main twitter posting code is based on what I learned from "The Reply to Mentions Bot" at https://realpython.com/twitter-bot-python-tweepy/#the-config-module

New version 2 Aug 2021 adds:

• run Apple Logo II code by using the {L} directive
• run Terrapin Logo PLUS code by using the {T} directive
• download and run any arbitrary DSK file from the web by including a URL (and nothing else) in the tweet. This doesn't work super reliably and not really what I want to do with the bot. I'll probably take it out in the future.

Dependencies. So many dependencies:

• A Twitter account, and API keys for it https://developer.twitter.com/en/products/twitter-api
• Tweepy. Specifically the fork that allows video uploads. tweepy/tweepy#1414 They plan on folding that feature into the main program but as of this writing, haven't.
• LinApple, an Apple II emulator for linux. https://github.com/linappleii/linapple. I'm using a version that's built for Raspberry Pi. Maybe it's this: https://github.com/dabonetn/linapple-pie or maybe it's this: https://wiki.reactivemicro.com/Linapple_Raspberrypi shrug
• The command-line version of AppleCommander. At http://applecommander.github.io and https://github.com/AppleCommander/AppleCommander
• Java, to run AppleCommander
• bastoken, a tokenizer for AppleSoft BASIC created specially for this project by KrisKennaway: https://github.com/KrisKennaway/bastoken. The bot expects it to be named tokenize.py.
• ffmpeg, for processing video files: https://ffmpeg.org
• in the assets/ directory: an Apple II DOS 3.3 DSK image confugured to run HELLO at boot, called DOS33FRESHIE.dsk
• in the assets/ directory: a blank ProDOS DSK image, called PRODOSFRESHIE.dsk
• in the assets/ directory, Apple Logo II disk image called apple_logo_ii.dsk. Not provided here because copyright, but it's not hard to find.
• in the assets/ directory, Terrapin Logo PLUS v1.2 Language Disk, named Terrapin1.dsk, and the Utilities Disk, named Terrapin2.dsk. Also not provided here because copyright.
• an X Virtual Frame Buffer running on display 98 (/usr/bin/Xvfb :98 -ac -screen 0 1024x768x24)

appleiibot's Issues

readme.md copy-paste error

Wonderful project Kay. I don't want to sully it with my name from a pull request for something so dinky: The readme says it is an "Atari 8-bit Twitter bot".

Character encoding limitations of Pyton

Somewhere in the execution chain (I suspect when opening a text file) data is being forced to ASCII and otherwise useful characters (upper-ascii, unicode, etc) are discarded. Because of this, it is not possible to post pre-tokenized or other kinds of encoded data and we have a pretty hard limitation of 140 characters to get past this. If this bot is using Python 2, then it is very likely that the default behavior of Python is responsible for this, as it assumes Ascii in a lot of places. If the bot is Python 3, then I don't have a reasonable explanation since I don't see normalize being called explicitly anywhere in the bot or the tokenizer code. At any rate, explicitly specifying UTF-8 as the encoding type when opening text files could be one possible way to solve this limitation. It is possible that the tokenizer would also need to be updated to truncate hi-bytes or do something else clever with them. Also, using unicode symbols could reduce the number of characters needed for commands, e.g. 🟢 for green, 🟠 for amber,
🐢for logo, etc. Of course if raw binary were also possible then maybe ⊛ for assembled data?