Adrian Cole

ps we still prefer route as path can have variables. can of course special case safe bets like root path. learning cardinality post factum is something fancy someone could do...

Ramon's opinion is to intentionally make high cardinality span names and have the backend ignore them (reassemble from attributes I guess) Is there anyone else in favour of this approach?

> > @adriancole one of scenarios for > tracestate requires in-process exposure and modification. Some systems want > to keep (and update) legacy ids at least for a while. assuming...

PS I think reserializing the tracestate into a string for every intermediate span will be unnecessary overhead for a lot of folks. We already imply a lot of overhead to...

FYI amazon call trusting upstream "pass-through" This is also the same default behavior as spring cloud sleuth http://docs.aws.amazon.com/lambda/latest/dg/lambda-x-ray.html

fyi you can look at brave which has a design including these features and not requiring inheritance. context has a list of "extra" which propagation plugins can add to. there...

> > @adriancole Could you provide an example > to show how to use the extendable mechanism? Such as exact context > key-value, notification when creating a span or inject/extract...

actually I should be clear I did find one thing close, just it was a bridge too far for what I needed. For the use cases we had.. it was...

in zipkin, there's an endpoint which includes a service name, which is a label, and network context. https://github.com/openzipkin/zipkin-api/blob/master/zipkin.proto#L164 There have been historical chats about how to represent activity of transparent...

@SergeyKanzhelev I suppose what I mean is a name, low cardinality aggregatable service name, so for example, not a generated IP based thing or UUID. Ex. if my service is...