Christian Neumüller

The solution used previously was that each vendor uses their own header name. The problem of overriding each other came to be exactly because we standardized on a single header...

In the end, if you can avoid it, you should not have multiple agents trying to do the same thing in the same process (same webpage).

> It also goes against the whole idea of the TraceContext standard that is designed to make multiple vendors work together. My impression is rather that tracecontext makes multiple tracing...

It's not a disappointing argument. This is part of the specification, one of the main uses of a specification is to have behavior that people can rely on. There is...

You can't bring that argument against any change of course, but "service" is a very fundamental concept and it's marked as stable, which means the OpenTelemetry project did commit to...

We need to also identify links created before span creation since some SDKs might already be adding them after creation, so the status of all links at present is "unknown...

You mention the span name, but you mean the description inside the span status, right?

> 2. propagate the random flag if a traceparent is received ~Isn't propagation of unknown flags already required by the previous W3C spec? Would still be good to verify I...

@adriangb The sampler intentionally does not receive the span ID, see https://github.com/open-telemetry/opentelemetry-specification/pull/621. But I wonder if the arguments from there are still usable, after all an unsampled span also receives...

Good point, I think you are indeed right here @AndersLindPedersen. The correct way to use this should be accessing the tracer in the lambda instead of extracting the tag beforehand:...