Issue 10156
Checklist dataset page, download: clarify whether download authorisation needs to be handled on portal side
10156
Reporter: omeyn
Type: SubTask
Summary: Checklist dataset page, download: clarify whether download authorisation needs to be handled on portal side
Priority: Major
Resolution: Fixed
Status: Closed
Created: 2011-11-07 16:58:25.366
Updated: 2013-08-29 14:46:11.962
Resolved: 2011-11-16 10:47:15.998
Author: omeyn@gbif.org
Created: 2011-11-07 16:58:51.292
Updated: 2011-11-07 16:58:51.292
Andrea Hahn Thu, 29 Sep at 8:09am
http://staging.gbif.org:8080/portal-web/dataset/1, block "Checklist preview": check with Dave whether we need to handle the case that a checklist owner does not want access to the full dataset (download). If yes, decide whether this would be our obligation, or that of the publisher (block access).
Andrea Hahn Mon, 17 Oct at 7:44am
Dave 17/10/11:
- yes, some checklist owners do not want to allow automatic download of their full checklist through the data portal (while browsing the checklist in the taxon browser is ok)
- as browsing the checklist is ok, and GBIF has the checklist data indexed and handles the download generation from there, download control for such cases needs to be part of the portal functionality
- it is fine to use the taxon browser module on the checklist dataset page, instead of the current taxon list fed from checklist metadata. This means that the taxon browser can be used to navigate to the root (or any other) taxon of the checklist in the species page context.
____
- there is no answer yet to the question of what would keep a user from downloading the checklist right at source (access point URL). Publishers need to be aware of that possibility. This might lead to a renewed request to handle access to protected URLs on the side of the indexer.
- open question: do we need to cater for checklist resources whose data are not yet accessible, but who do supply metadata (alternative representation from taxon browser)?
Markus Döring Mon, 17 Oct at 9:17am
Personally I think publishers should only themselves protect their data, i.e. the archive URL. If they don't do this, its publicly available and we should link to their URL for downloads, not repackage their data (although we might want to add nub urls and use interpreted data in the download - so we need to repackage them).
In regard to external metadata these should not show up as checklists, but as generic datasets. Only indexed checklists will appear as such
Andrea Hahn Tue, 18 Oct at 4:27am
Agreed about the non-repackaging of data.
Protected access: so far, GBIF has always been pointing at the policy of "free and open access", and refused to handle indexing from protected URLs. Needs further discussion at political level if this policy is to change. Agree that access would need to be protected at source if at all, as it is otherwise just an illusion anyway.
In practical terms this would mean for now:
- checklist download gets the source from registered access point
- download dialog only gives reminder re. correct citation etc., but does not manage download permissions
- the download dialog needs to handle the case of a failed download due to URL access restriction (feedback message to user)
- checklist publishers who do not want their checklist to be publicly available can be registered, but will presently not be indexed, unless GBIF changes their indexing policy about managing credential handling for indexing purposes. This implies that the taxon browser will not be available either for these datasets