signalupdates-bot

An unofficial bot that creates posts about new tags in Signal repositories in the corresponding beta feedback topics and sends update notices to Discord.

Running an instance

To run this bot, do the following:

Setup

1. Install Rust.
2. Follow the Cloudflare Workers Get started guide from the beginning up to, but not including, step 3.
3. Clone this repository using Git:

   git clone https://github.com/u32i64/signalupdates-bot

Configuration

General

1. Copy the wrangler.example.toml file to wrangler.toml.

2. In wrangler.toml, fill in the following:

   Field	Description
   account_id	Available in the Cloudflare dashboard.
   id for STATE binding in kv_namespaces	Create a KV namespace in the dashboard, then copy its ID. If you'll be deploying the staging variant as well, it is recommended to create a separate KV namespace for it.
   USER_ID	The bot's numeric user ID like 12345. You can find it by inspecting the HTML or JSON of any of the bot's posts. It is used to find the bot's posts when confirming post approval (which is done without the API key in case the post is returned even if it's not approved, so the yours property of posts can't be used).
   TOPIC_ID_FOR_SERVER_UPDATES	The topic's ID for posting server updates, for example 12345. TOPIC_ID_OVERRIDE overrides this one as well.
   TOPIC_ID_OVERRIDE	If you'd like all of the bot's posts to go to a single topic, set this variable to the topic's ID, for example 12345. Otherwise, leave it empty.
   DRY_RUN	If you'd like the bot to skip actually posting to Discourse, but otherwise do everything else, including modifying the state (with dummy post numbers), set this to true. Otherwise, leave it empty.
   ENABLED_PLATFORMS	Specify first letters of platform names you want to enable. For example, specify said to enable all supported platforms (Server, Android, iOS, and Desktop).
   DISCORD_UPDATES_MENTION_ROLE	Role ID to mention about new versions in Discord.
   DISCORD_SERVER_UPDATES_MENTION_ROLE	Role ID to mention about new Server versions in Discord.
   DISCORD_ERRORS_MENTION_ROLE	Role ID to mention about errors in Discord.
   ACCESS_TOKEN	Token that can be used to run the bot on demand or tell the latest Android Firebase version to it via API.

3. In the KV namespace(s) you created, manually create a key-value pair with the key state and a value like:

   {
       "android": {
           "last_posted_tag_previous_release": { "name": "v1.2.3" },
           "last_posted_tag": { "name": "v1.3.0" }
       },
       "ios": {
           "last_posted_tag_previous_release": { "name": "1.2.0.4-beta" },
           "last_posted_tag": { "name": "1.3.0.4-beta" }
       },
       "desktop": {
           "last_posted_tag_previous_release": { "name": "v1.2.0-beta.1" },
           "last_posted_tag": { "name": "v1.3.0-beta.1" }
       },
       "server": {
           "last_posted_tag_previous_release": { "name": "v11.99.0" },
           "last_posted_tag": { "name": "v12.6.0" }
       }
   }

   Adjust the value accordingly if you'd like to skip posting many old versions after starting the bot.

   You can also add other values used in PlatformState (see src/state.rs), but this is not required, as default values will be used automatically.

Discourse

Configure the bot's access to Discourse.

1. Create a new user for the bot in the Discourse instance, if necessary.

2. Log in as that user and create a Discourse user API key, for example using this CLI tool: KengoTODA/discourse-api-key-generator.

   To use this tool, you will need to install NPM, specifically you need the npx command from it.

3. In the folder where you have the bot's code, run:

   wrangler secret put DISCOURSE_API_KEY -e production

   When prompted, paste the API key you've created and press Enter.

   If you'll be deploying the staging variant as well, run the same command with staging instead of production (and also do this for future commands below). Note that you might not be able to create two different API keys using the tool above (when you create a new one, the old one may be removed); you may have to use the same API key for both production and staging.

   Note: you may have to publish the bot once before running this command. To publish it, see the Deploying section below.

Discord

To send the bot's log to a Discord channel in case an error occurs, set up a webhook.

1. Create a webhook in a Discord server (guild) by following this article.

2. Once you have the webhook URL, configure the bot with it, similar to what you did with the Discourse API key:

   wrangler secret put DISCORD_WEBHOOK_URL -e production

For sending update notices themselves to Discord as well, set DISCORD_WEBHOOK_URL_UPDATES and DISCORD_WEBHOOK_URL_SERVER_UPDATES in the same way.

Deployment

Run the following command to deploy the bot:

wrangler publish -e production

The production variant is configured by default to run every 10 minutes. For the staging variant, you have to invoke it manually by visiting its URL (that looks like signalupdates-bot-staging.<your-workers-subdomain>.workers.dev/ACCESS_TOKEN/run).

Acknowledgements

Special thanks to all participants of the discussion in this topic and to @newuser for developing their version of the bot.

Disclaimer

This is an unofficial project. It is not affiliated with the Signal Technology Foundation or Signal Messenger, LLC.