totominc / discord-stream Goto Github PK

Actually the HTTP service doesn't support API calls when the JWT is revoked.

Since the HTTP service is using axios, it's possible to add an interceptor (on a 401 response error) to do this flow:

1. keep initial request config (to re-try it later)
2. generate a new JWT
3. re-try the initial request

In the src/commands/invite.ts, the client ID invite link is not dynamic.

Instead, it should point to the environment variable USER_ID.

Sometimes we can't be sure the bot have the required permissions for a specific voice-channel, and it leaves a generic error message for the user.

The bot actually need 2 permissions (CONNECT, SPEAK) from the user voice-channel (message.member.voiceChannel). We can check if the bot doesn't have one of these permissions (message.member.voiceChannel.permissions.has(...)) and throw a less generic error message.

Actually the logs folder is at discord-stream/packages/discord-bot/logs but instead it should be at discord-stream/logs.

Also fix the logger to don't put Error strings on the errors.log file.

Add a tslint config with tslint-config-airbnb preset shared for all packages.

Then add a lint command to be executed by CI.

Actually when the discord-bot is making API requests it uses a JWT, which lasts for 1 hour. The bot is meant to be runned continually without interruption, this is where we need to revoke a JWT otherwise the bot won't be able to update data (i.e. a guild prefix).

There is an interval already setup which runs every 30 minutes to generate a new JWT, but instead we should use Axios's interceptors on a 401 failed response since it's the proper way to do it.

We should not pass any options when casually using the HTTP service. It should have better defaults such as Content-Type: application/json or automatically retrieve JWT (see #25).

The devDependency lerna was updated from 3.9.0 to 3.9.1.

🚨 View failing branch.

This version is covered by your current version range and after updating it in your project the build failed.

lerna is a devDependency of this project. It might not break your production code or affect downstream projects, but probably breaks your build or test tools, which may prevent deploying or publishing.

Your Greenkeeper Bot 🌴

Native promises are known to be slower (because they are not written in C, but instead in JS) and doesn't provide a lot of features.

Bluebird is a must have for any Node back-end, it is much more faster and use less memory. It has also cool features like promisify callbacks functions and built-in warnings and full stack traces (!).

When the bot already have queued tracks, if you try to add a track and it fails (returns an empty array of tracks or an error) the bot will leave the channel while speaking and having tracks queued.

For a potential fix, see this condition in the play.ts command

In the rich-embed response of a np command, it should add the current time of the track that is played.

When playing YouTube videos, the bot randomly don't play entirely YouTube videos. It doesn't have any problem when using other providers such as Soundcloud.

The issue is most probably related to the node-ytdl-core library since that's what we are using internally for video metadata & audio fetch/buffer.

An issue have been opened and is currently discussed on node-ytdl-core repo.

Don't use the inline parameter of a rich-embed as a fix.