Find out if icoscp_core can easily replace these next foundational points in the current icoscp 0.1.19:
resourcification (instead of envrification)
data & metadata access
http authenticated access vs jupyter open access
Things to address:
Should the validity of a PID be verified at the icoscp_core level or at the icoscp level?
From the landing page of a data object someone can grab metadata in various formats (json, ttl, xml, iso19115). Should this functionality be implemented in the icoscp_core library or rather provide the code for such conversions?
Deprecate dobj.id or Dobj.dobj properties. Both of these return the dobj. The dobj.id might cause confusion with the built-in id() function.
Comparing the "old way" (get request via the icoscp lib) and the "new way" (http request via the icoscp_core lib), it seems like class attribute coverageGeo is missing from the declaration of DataObject class in metacore.py
What about stilt authentication (which goes through atmo access)?
Find out if
icoscp_core
can easily replace these next foundational points in the currenticoscp 0.1.19
:Things to address:
coverageGeo
is missing from the declaration of DataObject class in metacore.py