Adrian Cole

@carlosalberto so this is still an issue.. I can see you last explained yourself are you waiting for me or someone to say go? I just recently had to do...

I understand how deprecations on api signatures work. I don't understand how one would accomplish a deprecated behavior like this: "Finishing Span upon Scope.close() has been deprecated." What does this...

Initial is that you should very quickly release a v0.33.0 which removes the deprecated methods entirely. I'll add some feedback based on thoughts in the diff https://github.com/opentracing/opentracing-java/compare/v0.31.0...v0.32.0 * `ScopeManager#active()` --...

@felixbarny I'll comment.. most of the bias is being least weird especially in apis meant to be like standard libraries. > Pros: Adding an MDC entry where the value is...

to predict a question on " It is much cheaper to add nothing" by anyone, not just @felixbarny do your own benchmark, and choose bench checking null vs invoking MDC...

Main concern summed up is that the current approach is the last thing we'd want to use for anything deemed common-case, as it is flawed and hard to spot how...

My concrete suggestion: rip out the the magic finisher part and keep the ability to implement scoping. Make the scoping thing like others where idiomatic in language (census, grpc, finagle,...

I was looking around on github to try and grok anyone's real world experience with ActiveSpan as it is defined today. What I found was that few tracers support it,...

@bhs I think we should consider desired practice at google if we are citing practice at google. census is being developed without this style, by folks at google. Specifically, I...

concretely, in census, there's a convenience type to close a span and a scope simultaneously. This is a lot different than counting calls to close and assuming a certain count...