I'll give them the benefit of the doubt and assume that's calculated 1 day in arrears... I actually only noticed they were down because we had an increase in our own app's errors which turned out to be our hipchat webhooks failing with 503s.

Either way I hope we get a root-cause analysis later, would be interesting to know what could take them completely down for 5+ hours. Presumably not a 'straightforward' bad deployment that could be rolled back.

They addressed it on twitter and said it only changes once an incident is resolved but they will look at possibly changing that.

I would be interested in a low level comparison of both HipChat/Slack's architecture and technology stack.