savetz / appleiibot Goto Github PK

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?

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".

I am aware of how Elon blew up Twitter, so I totally agree the bot there stopped.

A long term feature request (in the sense "in case anyone ever sends a pull request or you find time to investigate how this can be done") is to migrate this bot to Mastodon.

Since this bot was inspired by the BBC Micro one which already has migrated to Mastodon here is a link to its repository: https://github.com/8bitkick/BBCMicroBot