Major API changes — New Databento Binary Encoding (dbn) and timeseries.get_range behavior
Dear Databento API users,
Two API breaking changes will go into effect on March 3, 2023 at 5 PM EST. These changes are necessary to improve the functionality of our API and ensure that we continue to provide the best possible service to our users.
Change 1: Migration of Databento Binary Encoding from its legacy format to a newer format (from ‘dbz’ to ‘dbn’)
- The new ‘dbn’ encoding is nearly identical as ‘dbz’, but the two are not compatible.
- All prior code and client library functions that depend on ‘dbz’ will cease to work.
- The rename is made to better reflect the nature of the encoding – it no longer requires compression with Zstandard, as is expected to support low latency requirements for our upcoming live data service.
- Our documentation has updated to reflect the new format and presentation protocol.
Change 2: Renaming /timeseries.stream to /timeseries.get_range:
- This change is being made to better reflect the functionality of the endpoint, which allows you to get a range of data from a time series.
- The /timeseries.stream endpoint will be deprecated on March 3, 2023 and removed after June 2, 2023. We strongly recommend updating your code to use the new /timeseries.get_range endpoint as soon as possible after March 3, 2023.
What you need to do
1. Review our API documentation to understand the changes and how they will affect your integration. Release notes can be found at https://docs.databento.com/release-notes.
2. Update your code to use the new ‘dbn’ encoding and the new /timeseries.get_range endpoint. If you are using either of our official Python or C++ client libraries, you must update to the latest release on March 3, 2023.
3. If you have old ‘dbz’ files:
- You can still read these files with prior versions of our Python and C++ client libraries before March 3, 2023.
- We have released a dbn-cli tool that can help you convert your ‘dbz’ files to ‘dbn.zst’ files using the new encoding.
- You may also request through chat support for us to re-run your old batch jobs with ‘dbz’ encoding to generate equivalent ‘dbn.zst’ files, at no additional cost.
If you have any questions or need assistance with updating your integration, please contact us. We’re here to help and are committed to making this transition as smooth as possible for our users!
Best regards,
The Databento Team