spec-tacles / gateway Goto Github PK

How does the gateway handle this, if at all? Discord bots generally need information about the relevant guilds, channels, and roles, in order to do the appropriate permission checking. With a broker like this, it's necessary to store this information in an external cache (eg. Redis) to avoid losing the information upon the consumer reconnecting, since it's only provided by Discord at shard connect time (as part of the GUILD_CREATE events). Is there a solution for this in the Spectacles ecosystem at the moment?

Like there will be cache on redis about each server member and cache will be updated by discord events like guildmember add/remove/update - Which can allow us to bypass the the special intent feature introduce by discord as of we just fetch data from our cache and get any member details or there presence data ;)

The currently used ZSTD compression has been causing issues parsing insanely large (but essential) data from the web-socket. My bot has a guild with more than 400 channels and complex permission setup. The server data received via guildCreate event triggered on Ready was being trimmed no matter what.

• I tried raising the read and write limits of the websocket. Didn't resolve the issue
• Tried switching to accepting zlib encoding and changed the compression to zlib algorithm and this resolved the issue.

Perhaps an option in the gateway.toml to select the compression algorithm? Or a complete switch to zlib? Any of the above, I'll be more than happy to open a pull request with the changes.

The gateway should log on stderr and leave stdout for JSON output by the gateway.

It's fairly annoying to use a config file with Docker (either with volumes or a custom image with the config copied in), so we should support config through environment variables so users can configure the gateway directly from CLI or Docker Compose file.

I'm interested in using this for my bot, and making my own library for it that simulates JDA.
Problem is, where's the documentation?

Just a simple listing of what the spec is relating to RabbitMQ would go a long way.

If possible, I'd love to see support for clustering for Redis by setting an array of URLs on the REDIS_URL option.
I noticed that radix supports clustering, so to my understanding it should be possible.

I typed out a message join voice channel tried every events as mentioned but that didn't worked

We just got approved for GitHub Actions, so we should add actions to automatically build and publish releases.

resume_gateway_url has been out for websocket connections to use for a while now, do you plan on adding support for it?