My fix for issue #312995 addresses the case where -scheduleNextUpdateCheck is run anywhere other than main thread. It also properly clears out the checkTimer when invalidated, regardless of the originating thread. This guarantees that a background thread never gets the timer accidentally, and that we retain the timer (rather than relying on the runloop's +1).
My fix for issue #312995 addresses the case where -scheduleNextUp dateCheck is run anywhere other than main thread. It also properly clears out the checkTimer when invalidated, regardless of the originating thread. This guarantees that a background thread never gets the timer accidentally, and that we retain the timer (rather than relying on the runloop's +1).