Christian Neumüller

Links do not solve the problems discussed here (e.g. only the parent having the info whether it waits for the result, hence whether this is a sync/async call).

I have the feeling there are not enough people who want this strongly enough. If you want to get this rolling, you should probably make a spec PR or OTEP...

This is intentional. Maybe we should add a note in the spec about this?

This was discussed numerous times. Nested attributes make it harder for backends to process (search, store, ...) them. Issue #376 is still open in fact, I think this is where...

This would be the kind of the opposite direction of #2056, wouldn't it? I could imagine renaming to http.url_scheme (or http.uri_scheme) but I don't think there is a real problem...

You did not mention "Go" anywhere, but it seems you are actually benchmarking only the Go code for parsing/writing the protocol. The protocol will have different performance depending on the...

See also issue on HTTP app reporting (or per span service resource depending how you look at it): https://github.com/open-telemetry/opentelemetry-specification/issues/335 Also related: https://github.com/open-telemetry/opentelemetry-specification/issues/274 "Allow resources as span attributes"

Please give this PR a more descriptive title. I think the "inconsistency" issue has been explained. It seems to me this is not solving an inconsistency but proposing a new...

> I don't believe that this PR is identifying any new functionality, it is attempting to properly document what Attribute support is today. In practice, you are proposing a new...

What should be "clarified" then, is that Log Attributes are not the general attribute type. You have a clear definition of a general "Attribute" here which does not allow nesting....