Post 1850559 by TheHottestPotato@cybre.space
 (DIR) More posts by TheHottestPotato@cybre.space
 (DIR) Post #1850559 by TheHottestPotato@cybre.space
       2018-12-09T03:59:28Z
       
       0 likes, 1 repeats
       
       The decision in #activitypub that URL's themselves are the universally unique id's for all objects is extremely difficult to marry to a db system in an implementation without that implementation ending up feeling weird and messy. Like, do I just store the information that I need to know to check against _my_ urls and have separate tables for external statuses? or do I store a lot of redundant internal data so that I can have homogeneous table reads.