1.0.0: Cloud Hosting

With the core feature set where I wanted it as a minimum, I thought it was time to concentrate on Cloud Hosting for free. After experimenting with several options Oracles Cloud (OCI) won out as it seems to offer a tier that will give you always free hosting

You still need to clone the repo so that you can locally make any required changes to the config before starting the Bot for the first time, but once that's done and you've created the required OCI infrastructure (all documented) you can use the deploy script to pull a ready build image to the remote host and upload the config.

Updates should then simply involve running the deploy script again to update the image

0.9.8_beta: AFKMonitor, mod commands and themes!

• AFK Monitor with Auto-Removal for inactive DJs
• mod command with listUsers, remove and skip options so the bot can carry out some Mod tasks on your behalf, controlled via PMs
• owner-settable {theme} and intelligent {readTheme} (only outputs if theme contains something) tokens;
• per-user customisable welcome messages and images with the new !editWelcome command
• say command used via PM so you can make your bot say anything you want

0.9.7_beta: Non-Standard character substitutions for ML

Non-standard ascii characters used in some names, for users and hangouts, can become garbled in ML responses. This fix substitutes in standard ascii characters, and adds the charmap command allows users to add new characters as needed

0.9.6_beta: SQLite database and significant ML model changes

We're now creating an simple SQLite database to store, initially, DJ and song data. This is currently used to generate a token {last5plays} that can be sent to the ML Functions or used elsewhere in messages

Following the changes to the Google Gemini API limits on their Gemini models, the codebase has been rewritten to use the Gemma models instead, which have far higher daily thresholds for free use. However these are also simpler models so the prompts need to be far more explicit in their requirements.

Details on this in provided in the migration doc for this version

0.9.1_beta: Permissions fixes and new chatCommand sub-command

• Fixed an issue where coOwners couldn't use commands requiring the OWNER permission
• added a 'list' sub-command to the new chatCommand so you can see what's been created

0.9.0_beta: Dynamic chat commands

BREAKING CHANGE. CHECK THE DOCS AND UPGRADE NOTES IN THE CHANGELOG

Dynamic Chat Commands have arrived in V3.

With Dynamic chat commands anyone with Moderator or above can create a command and attach messages and gifs to it. Anyone can then use that command and a random message & pic from the selection is dropped into the chat.

Want to say hello regularly, create /hello and give it a couple of messages and nice pics

Now obviously images age out, which is why there's also an image validator that you can use to check that all your images still work, and delete any that don't

Also update the CHANGELOG as it was getting too big

0.8.5_beta: AI Enhancements, triggers and tokens...oh my!

• The "AI" system has been enhanced using common system instructions that are passed with every request to avoid repetition in individual commands
  • it's spilt into MLPersonality and MLInstructions. User one to describe the personality you want your Bot to have, and the other to direct it to specific sources of data, or to give other general directions
• Triggers have been introduced so you can configure the bot to perform commands when specific events occur
• Tokens have been standardised, and a custom token system introduced. This would let you, for example, create a token called {location}, use that token in a multiple messages or questions, and only have to update the token if it changes, not all the messages that use it
• Improved Documentation: added a file called CONTROLLING_YOUR_BOT in the docs folder describing how to use tokens, triggers, commands and features to customise your Bot

0.8.1_beta: Improved Edit Commands

Improved the edit command...use
edit list to see the available editable templates
edit show {templateName} to see the current contents of a template

0.8.0_beta: ML commands

Multiple ML Commands using Google Gemini added including,

• popfacts: displays 3 random facts about the current track
• whatyear: tells you when the playing track was released
• meaning: tells you what the lyrics of a song mean
• band: give you some background and history of the band playing

Note: the data returned is all from Google Gemini, so may not be entirely correct. Always take information returned by "AI" systems with a pinch of salt!

Update process

• The .env_example and data.json_example files contain all the new info
• add the following to your .env file and then follow the new section in the setup document for instructions on how to obtain your Google API key
• The bot will run without the key, but the new ML commands won't function

# Google AI / Gemini API key for machine learning features
googleAIKey=paste-your-google-ai-api-key-here

• Update your data.json file using data.json_example
• add the two new sections "mlQuestions" and "editableMessages" retaining your existing messages

0.7.0_alpha: Command & Feature control plus song announcements

UPGRADE NOTE
New persistent data exists in the data.json file. If you're upgrading from a previous version ensure that you copy the new elements out of the data.json_example file

New features

• Now Playing songs announced
• Just Played announcements made
• Features and Commands can be enabled and disabled using the new feature and command commands
• Now Playing message template can be edited using the editNowPlaying command
• Template changes and feature/commadn statuses are maintained between restarts using the data.json file