Adrian Cole
Adrian Cole
ps I'm not sure what the terms mean (pure wasm/universal wasm) but I would highly suggest having WASI work with whatever the lib attachments are. We have a comparison test...
sounds good to me. @dmvolod any objections?
@pannous agree on the dead thing, though hopefully runtimes don't accumulate so fast that marking dead ones manually becomes too much work ;)
one health signal troubling to me is where last commit is recent, but last version tag ancient. it is hard on the ecosystem this way, though forgivable for the first...
fyi I published this local and here are the differences, assuming you don't use the armeria-bom and just look at indirect deps from armeria-brave. I tried to match the pre-reporter...
Note: if 1.27.1 works at LINE than we actually don't need to pin to reporter 2.x. Basically if 1.27.1 works, it could be as simple as adding a zipkin core...
SUMMARY: My suspicion is that this is more about a "convenience dep" on `io.zipkin.zipkin2:zipkin` rather than specifically about reporter 2.x vs 3.x. If it was about zipkin-reporter 2.x vs 3.x,...
@jrhee17 is there any way to verify this at LINE with one of the apps that struggled or a similar reproducer? I want to make sure we get this right..
@jrhee17 so the thing is that it is possible to use zipkin2/reporter/brave/ZipkinSpanHandler without even the zipkin core library present. Because the input type is brave's MutableSpan. So, in this case...
actually zipkin2/reporter/brave/ZipkinSpanHandler needs zipkin core jar, I can see it does. The async one does not. I'll keep the zipkin core dep.