The OTel Ecosystem Explorer provides a website that allows users to explore the different telemetry that is available.
It currently only contains data for the Java Agent and the Collector - .NET does not feature: open-telemetry/opentelemetry-ecosystem-explorer#101.
Some initial work is happening in the ecosystem explorer to enumerator the available NuGet packages we provide, but that provides no insight as to the Semantic Conventions/versions/spans/metrics/attributes those packages provide.
If we were to include machine-readable metadata (which would also benefit AI Agents) in both this repository and open-telemetry/opentelemetry-dotnet-contrib for the packages we publish, this could be consumed by the Ecosystem Explorer and other tooling to provide a canonical expression of the data available. Being stored in the repository also allows it to be addressed by version using Git tags.
@matt-hensley proposed that the data format be compatible with weaver to allow it to be generated/composed with other the Semantic Conventions. A prototype of a possible approach doing so can be found in open-telemetry/opentelemetry-dotnet-contrib#4381.
One we have an agreed approach for where to store the data and in what format and the infrastructure in place, we could crowd-source (or use AI tooling) to generate the initial data on a per-package basis.
Over time this would then become part of BAU of adding and maintaining the instrumentation, allowing us to easily document the instrumentation we provide via the packages we maintaing.
The OTel Ecosystem Explorer provides a website that allows users to explore the different telemetry that is available.
It currently only contains data for the Java Agent and the Collector - .NET does not feature: open-telemetry/opentelemetry-ecosystem-explorer#101.
Some initial work is happening in the ecosystem explorer to enumerator the available NuGet packages we provide, but that provides no insight as to the Semantic Conventions/versions/spans/metrics/attributes those packages provide.
If we were to include machine-readable metadata (which would also benefit AI Agents) in both this repository and open-telemetry/opentelemetry-dotnet-contrib for the packages we publish, this could be consumed by the Ecosystem Explorer and other tooling to provide a canonical expression of the data available. Being stored in the repository also allows it to be addressed by version using Git tags.
@matt-hensley proposed that the data format be compatible with weaver to allow it to be generated/composed with other the Semantic Conventions. A prototype of a possible approach doing so can be found in open-telemetry/opentelemetry-dotnet-contrib#4381.
One we have an agreed approach for where to store the data and in what format and the infrastructure in place, we could crowd-source (or use AI tooling) to generate the initial data on a per-package basis.
Over time this would then become part of BAU of adding and maintaining the instrumentation, allowing us to easily document the instrumentation we provide via the packages we maintaing.