openEO

New releases for openEO API and openEO Processes, focus on vector data

Written on May 25, 2023 by Matthias Mohr.

Today, we have released the new version 1.2.0 of the openEO API. At the same time, we've also released a 2.0.0 release candidate for the openEO processes.

The focus of these releases is to implement vector data (cube) support in openEO so that the focus is not strongly on raster data any longer.

Back-ends and clients are advised to check their implementations against the new specifications and update accordingly.

Below you can find details about the individual releases:

# openEO API v1.2.0

We've collected improvements and useful new features for two years. It contains clarifications, new extensions (for orders and federation), vector data cubes, STAC (API) updates, alignment with OGC APIs, more link relation types, improved batch job results and logs, and other minor improvements. This API release is required to be able to release openEO processes in version 2.0.0-rc.1 due to the changes from the "raster-cube" to "datacube" subtype and the related changes in the process schema.

Please consult the changelog (opens new window) for the changes and additions that have been made to the API.

# openEO Processes v2.0.0-rc.1

AsWe've collected improvements and useful new processes for over a year. The processes incorporate the long-awaited changes for vector data cubes. Unfortunately, this involves some breaking changes (e.g., the new datacube datatype) and thus we are moving from 1.x to 2.x. Due to the amount of (breaking) changes, we propose the changes to the community as a release candidate first so that we can evaluate whether it all works as expected. We'll then cut the final release once some implementations confirm that the changes work as planned. Feedback is more than welcome and we'll closely monitor the GitHub issue tracker (opens new window) for your issues. Please feel encoruaged to post your feedback there.

This release includes (for the first time) a couple of breaking changes. Therefore, we'll highlight the breaking changes below. Please consult the changelog (opens new window) for a full list of changes and additions.

  • Added better support for labeled arrays. Labels are not discarded in all cases anymore. Affected processes:
    • array_append
    • array_concat
    • array_modify
  • array_modify: Change the default value for length from 1 to 0. #312 (opens new window)
  • aggregate_temporal, filter_temporal, load_collection and load_result/load_stac:
  • inspect: The parameter message has been moved to be the second argument. #369 (opens new window)
  • New definition for aggregate_spatial:
  • Updated the processes based on the subtypes raster-cube or vector-cube to work with the subtype datacube instead. #68 (opens new window)
  • sort and order: The ordering of ties is not defined anymore. #409 (opens new window)
  • quantiles: Parameter probabilities provided as array must be in ascending order. #297 (opens new window)
  • fit_curve and predict_curve: Heavily modified specifications. fit_curve works on arrays instead of data cubes, predict_curve doesn't support gap filling anymore, clarify no-data handling, ... #425 (opens new window)
  • climatological_normal: The climatology_period parameter accepts an array of integers instead of strings. #331 (opens new window)
  • between: Support for temporal comparison. Use date_between instead. #331 (opens new window)
  • Deprecated GeometryCollections are not supported any longer. #389 (opens new window)
  • Deprecated PROJ definitions for the CRS are not supported any longer.
  • The comparison processes eq, neq, lt, lte, gt, gte and array_contains:
    • Removed support for temporal comparison. Instead explicitly use date_difference.
    • Removed support for the input data types array and object. #208 (opens new window)
  • sort and order: Removed support for time-only values. #331 (opens new window)