Project is being Sunset

[MikeGoldsmith]

This project has outlived its useful lifespan, and we intend to deprecate it in the future. Honeycomb recommends using the official OpenTelemetry Python SDK, and using individual contrib components as required. The official OpenTelemetry Python SDK is stable and has grown to offer most of the ergonomic benefits that this distribution has historically offered.

If you are actively using this distribution and feel the official OpenTelemetry SDK does not meet your needs, please file an issue here with your use case.

The following links may be useful:

• Honeycomb OSS lifecycle and practices.
• Configuring the OpenTelemetry SDK for Python with Honeycomb.

[kmcquade]

I really like the local visualizations feature: https://docs.honeycomb.io/send-data/python/honeycomb-distribution/#visualizing-traces-locally

My biggest question - it lists these advantages of this package in the docs (copied and pasted below):

• more easily initialize OpenTelemetry for use with Honeycomb
• more easily add multi-span attributes
• more easily configure sampling

Do you have recommendations about how we can realize these benefits without necessarily using the Honeycomb distribution?

What is your timeline for deprecating this package, and will you be incrementing any new versions until then?

[MikeGoldsmith]

Hi @kmcquade - thanks for your feedback.

The official Python OpenTelemetry SDK has improved with regards to configuration and doesn't feel that much different from the Honeycomb distribution now. You can see our docs on setting up the otel SDK directly to work with Honeycomb here. This includes configuring sampling and setting the Honeycomb specific sampleRate attribute.

We're working contributing the baggage span processor to the otel-python-contrib project as that is something many users, including non-Honeycomb users, can benefit from.

We will be continuing to maintain this project while gathering feedback on sunsetting the project and anticipate it will be a multi-month process. New versions of the package will be published as required.

[kmcquade]

Okay awesome. Really appreciate that the team will continue updating the package and sunsetting it over a multiple month process. and I totally agree that all users can benefit from the baggage span processor.

Thanks!

[etimberg]

One thing that I found very helpful with this package was the automated Resource creation which has to be done manually when using gunicorn per https://opentelemetry-python.readthedocs.io/en/latest/examples/fork-process-model/README.html#working-with-fork-process-models