It seems possible to me that the Solr datestamp field may not always accurately reflect the date of last update, in the cases where e.g. an object has been removed from a narrative. In such a case, the lastUpdated date of the narrative would reflect today's date, but the datestamp of the object record will be set to the latest one of the lastUpdated dates of the object itself and all its related entities, which will not include the now-disconnected narrative. The Solr record will be updated OK, because the source_count will have been decremented by 1 and will no longer match, but the resulting Solr record's datestamp will not be updated to the current date, and this could hide some updates from harvesters.
It seems possible to me that the Solr
datestampfield may not always accurately reflect the date of last update, in the cases where e.g. an object has been removed from a narrative. In such a case, thelastUpdateddate of the narrative would reflect today's date, but thedatestampof the object record will be set to the latest one of thelastUpdateddates of the object itself and all its related entities, which will not include the now-disconnected narrative. The Solr record will be updated OK, because thesource_countwill have been decremented by 1 and will no longer match, but the resulting Solr record'sdatestampwill not be updated to the current date, and this could hide some updates from harvesters.