easy
easy
Chase pointer: Envoy discussion about this: https://github.com/envoyproxy/envoy/issues/7422
How important is the "abandon" use-case? Instead of abandon, could we satisfy this use-case by filtering out spans by name during export?
I've started work on this.
This is what I've got so far: https://github.com/g-easy/opencensus-cpp/tree/appveyor It gets surprisingly far but build is broken on the proto targets with underscores that we filter out on Travis, and test...
I'm running into `Cannot open compiler generated file: '': Invalid argument` - this is https://github.com/bazelbuild/bazel/issues/4149 (cl.exe can't handle long paths and gives a confusing error) My workaround for now is:...
We don't build everything and we don't test anything. It looks like isturdy was working on (thanks!) getting the gRPC example / integration to build but ran into a grpc...
#292 replaced Appveyor Bazel build with a CMake one. I'd like to see Bazel come back, and also for everything to be built and tested like it is on Travis/Linux.
Where does the StrictSpans idea come from? I can't find any references to this term.
This PR adds a new TraceOptions bit. Could you please open an issue at https://github.com/census-instrumentation/opencensus-specs and describe the motivating use case for "strict" mode. There are several implementations of OpenCensus...
Yes, doing this would distort the overall picture because the sampling would no longer be uniform. We can mitigate that by annotating traces as "sampled uniformly" versus "sampled because something...