Nic Munroe

Apologies in advance if I'm missing something and the following is dumb. In the case of receiving a span ID but no trace ID I can understand throwing stuff away...

> @nicmunroe on part 2: If the user can send trace id, they can also send a span id, right? Let me know if #4 (comment) doesn't cover that. @adriancole...

@adriancole ok, a few things are clicking for me now and I think I see where some of the disconnect on my part was: 1. Wingtips generates a new random...

Yeah in wingtips the server always separates the spans by taking the client's span ID as the server's parent span and creating a new span ID for the server span....

@yurishkuro I agree with you in principle that bad instrumentation is possible to detect proactively, but not every organization is mature enough or has built the proper tools and alerting...

We return X-B3-TraceId in response headers by default. Unfortunately there are some cases where a caller needs to know what the trace ID will be before it makes the call...

@adriancole could the B3 docs state that it's an application concern whether or not the trace ID is returned in the response? i.e. it's not part of the spec and...

@adriancole totally understood on wanting to avoid situations that generate timewasting questions, especially on something that doesn't need to be (or shouldn't be) part of the spec. I support whatever...

We just ran into this, and the panic in `context.Reset(...)` can be hit without it being a race condition at all - it's just a fundamental mismatch of `context` (or...

> At this point - routes (adding a route, note: there is no way to remove a route) or any of the echo fields is not meant to be mutated...