Ted Young

Ah, I see I am incorrect. I thought `apply` had similar issues to `bind` and prevented object caching.

Good idea. 👍

I'd also like to see this move forwards as an RC, so that tracers will binding to it and we can get more feedback. I like the idea of providing...

Hi Felix! I agree as well, on the broad point that SpanContext does not reveal enough of the span's context. Getting people to agree on what more should be exposed...

@yurishkuro I've been thinking about this, and so far can't come up with a better approach than your observer pattern. The only things I don't like about it: - it...

Yeah, I agree a contrib project is a good idea. These Observers/hooks are consumed by the tracer implementations, optionally supported, and aren't directly part of the OT-API, so contrib might...

@rbtcollins I'm sure we'll do more copy editing, but the docs are updated: https://github.com/opentracing/opentracing-java/tree/v0.31.0

I have one of these that is pretty straight forwards. Written for golang, then ported by @carlosalberto to java: Go version: https://github.com/tedsuo/ot-walkthrough-go Java version: https://github.com/carlosalberto/ot-walkthrough-java However, they are not quite...

Yes, I think you want something more like a patterns library or "The OpenTracing Cookbook" that helps teach people how to instrument applications, libraries, and execution frameworks.

@mabn, I wonder: if we can get SpanID and TraceID added to SpanContext, does that make Observers work well enough to accomplish MDC integration without a span wrapper?