Skip to end of metadata
Go to start of metadata

I. Introduction

1. At its meeting on 1 November 2011, the High-Level Group for the Modernization of Statistical Production and Services (HLG):

"emphasised the need for a global approach to statistical software sharing, and the need for software to be properly documented, including a clear label as to whether it is fully compliant with the HLG vision or not. The inventory maintained by the Sharing Advisory Board will be modified accordingly"

2. Objective criteria are therefore needed to determine compliance with the HLG vision, as well as a mechanism for applying them. This note proposes a set of suitable criteria.

II. Scope

3. Two types of software can be distinguished: commercial and non-commercial.

4. Commercial software includes all programmes produced by businesses seeking to make a profit. Note that commercial software may sometimes be free of charge, e.g. demonstration versions, readers or other cut-down software tools that can be used to try to persuade users to buy more comprehensive versions.

5. Non-commercial software includes all programmes produced by non-profit and government sector organisations, and all software developed by user communities on an "open source" basis. Note that not all non-commercial software is free of charge, in some cases license or other fees may be charged to cover administrative or development costs, or for user support or training.

6. The scope of the software inventory developed by the Sharing Advisory Board in cooperation with the Common Reference Architecture (CORA) project was non-commercial software developed by national or international statistical organisations. This scope will be widened to cover all non-commercial software according to the above definition, where it is used by one or more statistical organisation.

7. Commercial software remains out of scope due to the sensitivities around appearing to endorse particular commercial products or suppliers. However, a list of commercial software tools used by statistical organisations is proposed, purely as a reference tool.

III. Technical Criteria

8. To be considered compliant with the HLG vision, software must:

  • Support input and output of data (and metadata) in "open" formats, e.g. CSV, DDI, XML (including SDMX)
  • Be capable of being used in a "plug and play" architecture. The term 'plug and play' refers to an evolving approach in designing and implementing statistical software solutions, largely based on the "service-oriented architecture" model (SOA see http://en.wikipedia.org/wiki/Service-oriented_architecture). The key requirements for the HLG vision are:
    • Having clearly defined interfaces to run the software
    • Having clearly defined outputs from running the software - both during run-time and at the end of the process
    • Using the SOA concept of appropriate granularity in implementation, loosely based on the sub-process level of the GSBPM
    • Designing and implementing for re-use by others.
  • Comply with the Principles and Guidelines on Developing Multilingual Applications for Official Statistics


IV. Other Criteria

9. The software must:

  • Have detailed technical and end-user documentation available in English
  • Have a clearly stated licensing mechanism (e.g. open source, creative commons, annual fee etc.)
  • Be available for sharing by all official statistical organisations


V. Applying the Criteria

10. The "owners" of the software will have the primary responsibility for deciding whether software meets the above requirements. The Sharing Advisory Board will provide guidance on the application of the criteria, and will adjudicate in cases of disagreement. Software will be categorised as fully compliant, partially compliant or non-compliant.

VI. Reviewing Compliance

11. Entries in the software inventory should be reviewed annually. If there is new information affecting whether or not the software is compliant with the HLG vision, the software "owner" and the Sharing Advisory Board will review its categorisation.

  • No labels