Basic backup and recovery of the error queue, making it less likely that servers will lose errors in the event of a crash or restart (not perfect though).
Fully respect Discord rate limits
(Addresses #30)
Some notes:
cfc_err_forwarder_interval has been changed to cfc_err_forwarder_dedupe_duration. The setting is similar, but now has a more focused purpose, justifying a rename.
ErrorForwarder will not retry failed webhooks, it will simply log the failure and move on. My thinking is that servers typically have a reliable connection (the one exception being a message that's too long), so any issue causing requests to fail would likely cause all requests to fail. To that end, I'm not sure how useful retries will actually be.
This PR accomplishes two things:
(Addresses #30)
Some notes:
cfc_err_forwarder_interval
has been changed tocfc_err_forwarder_dedupe_duration
. The setting is similar, but now has a more focused purpose, justifying a rename.ErrorForwarder will not retry failed webhooks, it will simply log the failure and move on. My thinking is that servers typically have a reliable connection (the one exception being a message that's too long), so any issue causing requests to fail would likely cause all requests to fail. To that end, I'm not sure how useful retries will actually be.