I was wondering if there are any thoughts on being able to configure Flarum to support master/slave replication by sending all write requests to the master and all read requests to the slave. This would allow me to support having geographically spread Flarum for faster responses across the world. I'm already doing this with MediaWiki and it's working great. Our time to first paint has dropped from an average of 2 seconds in Europe to under 0.5 seconds for MediaWiki.
I'd like to do the same for Flarum. Granted, Flarum is much faster already and only needs API requests, but if the API response could potentially take half the time (or less) why not? The server and infrastructure is already there for MediaWiki, would just be piggy backing on that.
I'm now pairing MediaWiki and Flarum together and have SSO across them working perfectly. Users are able to log into MediaWiki and simply click the Forum link to be taken to Flarum and be logged in. If the user isn't logged in, they click on the 'Login' button within Flarum, and they get logged via MediaWiki.