Closed mbjones closed 6 years ago
Original Redmine Comment Author Name: ben leinfelder (ben leinfelder) Original Date: 2013-02-12T23:45:22Z
This is tricky. I know we decided a while back that we should manage the revision history separately for local vs. network objects, but in the old days when we used scope.docid.rev we implicitly considered revisions across their locations (so a newer version on the network does actually obsolete an older version locally even if it isn't yet saved locally).
I think we might have to manage revisions centrally so that we allow something on the network to obsolete something locally (and the other way around). There's another bug that might be related to this involving the search results: #5805
Original Redmine Comment Author Name: Jing Tao (Jing Tao) Original Date: 2013-02-13T02:05:43Z
Since the other way - saving locally first then saving both, works, we move this bug to 2.1
Original Redmine Comment Author Name: ben leinfelder (ben leinfelder) Original Date: 2013-03-05T01:10:48Z
Jing and I talked about a related issue today and decided that we CAN leave the
Original Redmine Comment Author Name: Jing Tao (Jing Tao) Original Date: 2013-03-05T23:07:27Z
Add the code to set the obsoletes element in the system metadata and the revision manager even though the obsoleted id doesn't exist locally. This is only for the local saving.
Original Redmine Comment Author Name: Redmine Admin (Redmine Admin) Original Date: 2013-03-27T21:31:55Z
Original Bugzilla ID was 5860
Author Name: Jing Tao (Jing Tao) Original Redmine Issue: 5860, https://projects.ecoinformatics.org/ecoinfo/issues/5860 Original Date: 2013-02-11 Original Assignee: Jing Tao
Here are the steps to produce the bug:
Create a data package with data object and save it to the network. It contains: resourceMap_foo.1, foo.1 and data.1
Update data.1 and save it to both network and local. It contains resourceMap-foo.2, foo.2 and data.2.
When we check the metadata for the resourceMap_foo.2, we found it has theresourceMap_foo.1 . Actually the local doesn't have resourceMap_foo.1.
Same thing to the foo.2 and data.2.