Merging IATI files

[timgdavies]

We can envisage a world in which:

• A large donor publishes IATI XML without ag-specific fields;
• Someone (possibly them) creates a file listing all the IATI activity ids and their additional ag-specific classifications

The challenge is how to combine this additional classification information with the original XML so that:

• (a) There is IATI XML file with all the information in;
• (b) That enriched file is discoverable through the IATI registry;
• (c) ideally, so that the original donor can bring the data back into their systems, if they want to.

In thinking about this we can to consider whether such a system can be:

• Evergreen - not needing maintenance and updates each time the standard updates;
• Adapted for other contexts - not just ag

[rolfkleef]

Some thoughts:

• Enriching data would make you a secondary publisher in IATI terms: we'd need to clarify how a data user can/should merge data from primary and secondary publishers. (Does data from a primary publisher take precedence? Even if it is known to be incomplete, erroneous, or misleading?)
• Becoming or staying a publisher in the IATI registry for this purpose might require some lobbying/negotiating, especially if enriching the data from an "adverserial point of view" leads to conflicting data?