[ckan-discuss] CKAN as a general-purpose cataloging tool?

[Allan Hollander]

Hi all -- I would like some feedback as whether it would be wise to use 
CKAN as a tool to catalog a heterogeneous set of resources, some of 
which are datasets in a strict sense and some of which are not. To 
elaborate, my information set includes a number of spatial and 
non-spatial datasets that all relate to environmental indicators on 
sustainability. The indicators themselves however need their own entry 
in the system, as does each member of a list of environmental issues to 
which these indicators in turn relate. There are other pieces to our 
information set but that gives the idea. In general, representing the 
linkages between these diverse resources will be important to our web 
application.

It seems based on my read of its architecture CKAN could be used for 
this task provided one changes its nomenclature. That is, everywhere 
where the term "dataset" is used in the interface replace (through 
template overrides) this text with "resource" and everywhere where 
"resource" is used replace this with say "resource component". The 
different broad categories of information sets (e.g. datasets, 
indicators, issues, bibliographic sources...) would be bundled together 
as CKAN groups. The relationships between each of these resources would 
be encoded using the package_relationships API (though I'd probably add 
a few more relationship types to the hardcoded list of these).

Does this sound like a feasible approach to using CKAN to build a portal 
for my information set? I would love to take advantage of all the 
beautiful UI work that has gone into CKAN 2.0, especially including the 
faceted browsing and the geospatial search.

Thanks,

Allan Hollander
Information Center for the Environment
University of California, Davis