Closed thegreystone closed 5 years ago
...you'd expect APM vendors to complain a bit more about this, as some vendors don't model this the same way as Zipkin and Jaeger.
The opentracing-0.32.0 branch currently shows what it would look like.
The spec has changed, and there is at least a spec change before we get the parent. Ergo, let's skip the parent for now, and get that later, whenever the spec is updated.
No point in waiting, as described in #14.
This would mean we can get at least get the spanID and traceID directly from the span. ParentID seems to be beyond reach still.