In a previous issue (#126) we expanded support to all available living atlases. We currently (v 1.5.0) support the GBIF name-matching service for some atlases, but no other services; I suspect this is a strange oversight to most users.
The first set of challenges to upgrading to GBIF are:
No GBIF API for field names, assertions etc (rgbif hard codes them)
Need for new translators in galah_filter as GBIF doesn't use solr
The 'predicates' system for GBIF downloads is analogous to our solr approach (albeit in JSON), but the 'online' GBIF occurrence API doesn't use them, potentially creating a mismatch between atlas_counts (online) and atlas_occurrences (offline)
Other problems may emerge once these are solved. However, the two systems are otherwise closely related, so it is conceptually straightforward to map current galah functions to GBIF.
In a previous issue (#126) we expanded support to all available living atlases. We currently (v 1.5.0) support the GBIF name-matching service for some atlases, but no other services; I suspect this is a strange oversight to most users.
The first set of challenges to upgrading to GBIF are:
rgbif
hard codes them)galah_filter
as GBIF doesn't usesolr
solr
approach (albeit in JSON), but the 'online' GBIF occurrence API doesn't use them, potentially creating a mismatch betweenatlas_counts
(online) andatlas_occurrences
(offline)Other problems may emerge once these are solved. However, the two systems are otherwise closely related, so it is conceptually straightforward to map current
galah
functions to GBIF.