Closed dyladan closed 1 year ago
I had been waiting for the spec->semconv transition to get back on this. I'll update this OTEP this week with feedback received.
I limited the scope of the proposal to just an attribute dictionary to address concerns raised by @lmolkova and others. I believe 99% of the value of the proposal is retained simply by allowing instrumentation authors to use attributes that are not already included in a semantic convention. If it goes well, maybe the idea will be extended to other areas of semantic conventions in the future.
@dyladan Here's a draft PR, proposing initial thoughts on how the yaml syntax for attribute definitions could look like.
https://github.com/open-telemetry/build-tools/pull/191
I'd appreciate your thoughts on that! Maybe we could even add this / similar yaml syntax spec as a potential example how it could look like into this proposal?
After discussion with @jsuereth we have decided to close this OTEP. A lot of what this OTEP was asking for is already implemented. Specifically, the Attributes Registry replaces the Attribute Dictionary mechanism described here.
We still believe that the core issue of lowering the barrier to create a new entry in the Attribute Dictionary is important, but believe that this OTEP is no longer the way to move that idea forward.
TODOs