"After decades of every vendor/project/application building their own implementation, we're finally going to have a commonly available implementation and standard from the OpenTelemetry project"
"To achieve 10-100 microseconds latency 99% of the time, you need to target 1% utilization of a core or less - 1% utilization of a core means 1% of the time that a new event comes in you're still dealing with the last event, so 1% of events are slowed"
