Open jqnatividad opened 10 years ago
Linux/OSX/BSD only I suspect, but writing a Fuse filesystem that uses the CKAN api seems entirely feasible and not too hard.
This wouldn't require any changes to CKAN itself, you'd just need ckanapi and fusepy
Yes @rossjones, FUSE would be the ideal way to do this. I've used Fuse and its quite powerful. Adding this functionality would increase usability in a number of ways:
And for people who prefer to dump files first, add metadata later (not me, honest), it could be v. useful :)
What if a CKAN Catalog can be mounted as a filesystem (NFS, Samba, etc.)? And then you can have pseudo-directories for groups and organizations? Won't it make it easier for data publishers and consumers to use CKAN beyond browsing the catalog? Won't it make it more straightforward to target CKAN for ETL scripts?