Bug #3006
openAdd LSID+GUID to data rows
0%
Description
The best way of tracking duplicates is unfortunately not our current system, but rather a GUID that is an LSID that has a GUID in it:
format: urn:lsid:cvs.bio.unc.edu:[tableName]:PK-[GUID]
length should be at least 150, probably go to full 255.
This would have to be override-able as sometimes newly created entities should map onto existing ones. This would enable each client entry tool to create unique IDs for data- without the need to update each client after adding to archive db.
This is far better than our current PK+bln_new flag. It would be a new series of fields (or one table?) that would create GUIDs and assign them to entities.
Updated by Michael Lee over 14 years ago
viewer version v201_p0730_mid3 now creates lsids for observation
Updated by Michael Lee about 14 years ago
This has been done for observation records, which are seen as critical for avoiding duplication. Other entities can wait.