When I saw the unexplainable failures today, I, too, couldn't help but think maybe it was finally time to take Travis out back. What with the reduced capacity and impending travis-ci.org shutdown (open source projects will continue to build, just on a travis-ci.com instance), now's probably as good a time as any. I haven't been a huge fan of how travis has been going since they were purchased. The thing that keeps getting me, though, is I know the github actions and ADO are hosted on essentially the same infra (Azure), so could have correlated outages - and outage avoidance (so we could still get some PR statuses when there was a cloud outage of some sort) was one of the only reasons we were using multiple providers to begin with. By dropping travis, we'll finally concede that we don't really care about that kinda uptime anymore. 🤷