SDMX 3.1 Standards. Section 1. Summary of Major Changes and New Functionality
- Contents
Revision History
Revision Date Contents DRAFT 1.0 December 2024 Draft release updated for SDMX 3.1 for public consultation 1.0 May 2025 Public release for SDMX 3.1 1. Overview
SDMX 3.1 is a minor revision to the SDMX 3.0 Standard which introduces a limited set of changes, which cover the following:
Information Model
- Support for Dataflows to reference a subset of Dimensions from a Data Structure Definition
- Simplification to Data Constraints
- Addition of Availability Constraints
Documentation
Registering Reference Metadata removed from documentation, to align with XML Registration object which is unable to reference a Metadata Provision, and REST API which is unable to query for registered reference metadata sources.
Breaking Changes
- Remove version property on Categorisation
- Removal of Advanced Release Calendar
Content of the Document
The remainder of this document contains a summary of the changes. More detailed information can be found in the SDMX 3.1 Technical Specifications, in particular:
Section 2 – Information Model
Section 5 – Registry Specification
Section 6 – Technical Notes
SDMX TWG GitHub for the REST API and the XML and JSON formats
2. Summary of Breaking Changes in 3.1
Version 3.1 introduces breaking changes, in the model in the following areas:
2.1 Removal of Advanced Release Calendar
The Data Constraint was simplified to become a cohesive structural artefact focussed solely on the restriction of values that can be reported. As part of this remodelling the Advanced Release Calendar was removed as it does not play a restrictive role.
Guidance for Implementors
Reference Metadata reported against a Dataflow and / or Provision Agreement can be used as an alternative to the Advanced Release Calendar.2.2 Removal of Version on Categorisation
The Categorisation structural artefact has no use case for undergoing version changes, the version was therefore removed.
Guidance for Implementors
If multiple versions of a Categorisation exist where the source or target differs, these Categorisations should be given a unique ID.3. Information Model
The changes to the information model support two distinct use cases:
- Horizontally complex Data Structure Definitions (DSDs)
- Additional cohesion to the Data Constraints model
These changes are described in below.
3.1 Horizontally Complex Data Structure Definitions
An explanation of this use case, with additional details can be found under Section 6 - Technical Notes, under heading ‘10.3 Dimension Constraint’.
The following changes to the model have been made to satisfy this use case.
Data Structure Definition
The Data Structure Definition has a new ‘evolving structure’ property, this is a Boolean property, when set to ‘true’ it indicates to the users of the DSD that new Dimensions may be added without the DSD undergoing a change to its major version number.
Dataflow
The Dataflow has a new Dimension Constraint property, which is used to reference a subset of Dimensions which it uses from the referenced DSD.
The Dimension Constraint fixes the Dimensions of the Dataflow enabling the referenced DSD to have new Dimensions added without the need to change the major version number.
The Dimension Constraint property is only required if the DSD sets the evolving structure property to true.
3.2 Constraint Cohesion
The Constraint model in SDMX 3.1 has been made more cohesive by separating the Data Constraint into two distinct structures; the Data Constraint which describes reporting restrictions, and the Availability Constraint, which describes data content from a data source. In SDMX 3.0 these distinctions were made using the ‘type’ property on a Data Constraint.
Data Constraint
The Data Constraint has had the ‘type’ property removed; the Data Constraint in SDMX 3.1 is always used to describe restrictions on content for data reporting purposes.
The attachment of the Data Constraint in SDMX 3.0 included Data Sources, in SDMX 3.1 these attachments have been removed as these are not relevant for restricting reported data.
Availability Constraint
The Availability Constraint is a new Structure introduced to describe data that exists, it is generated in the response to the Availability REST API. It is not a maintained structure, and as such has no maintenance agency, identity or version