Daniel (dB.) Doubrovkine

I think we're missing logging for cancelled tasks similar to slow query logging. Users will want to turn on something like "cancellable tasks logging" to enable dumping the query body...

@debora-ito I think it's the latter: a feature to support compression (#3286), and chunking (this request) per request. We're not talking to S3, but Amazon OpenSearch, so not sure what...

This worked for me. But seriously ...

Check out https://github.com/dblock/danger-changelog ;)

That *is* danger-changelog @mbuhot in many ways.

This is good, but I think we can do better. What do you think of being able to write this: ```ruby config.connection do |connection, options| connection.use ::Slack::Web::Faraday::Response::StoreScopes, options end ```...

> @dblock How would your first option work exactly? Would the block be stored in the config as a proc, and then called with the connection instance during the `::Faraday::Connection.new(endpoint,...

I have production services on https://www.playplay.io with many hundreds using https://github.com/slack-ruby/slack-ruby-bot-server which uses this library. The biggest issue becomes memory and startup time: ~1/3s to start (https://github.com/slack-ruby/slack-ruby-client/issues/87 needs fixing) and...

:+1: on exposing the entire thing, maybe like Hyperclient that supports `config.connection do ...`? https://github.com/codegram/hyperclient. Maybe try to contribute that?

It's seems not recommended because a lot of Faraday construction is middleware and you often can't modify the connection after-fact from what I observed in a predictable and useful way....