Brutal.

When Microsoft acquired GitHub.

@ironicbadger Could anyone explain to me, how is this possible?
I would imagine they would keep things running on the original hardware etc. which I wouldn't expect to fluctuate like this.

@Jourei @ironicbadger Yeah, this is likely just a bad graph. Before vs after the change was likely a change in either the definition of, or the measuring of "down time", and the two end up not comparable.

I suspect the data went from tracking as "most/key things working=up" to "anything not working=not up"