* [2011-02-14 15:03:19 -0500] Kingsley Idehen <kide...@openlinksw.com> écrit:
] If you can graph graph IRIs and scope them to your applications needs, ] what's wrong with that? You don't have to set up multiple ports or ] anything, you just have Graphs that are accessible to users via their ] SQL or WebID style identifiers. What if there are very many (e.g. millions of) named graphs? It seems to me that creating and maintaining the ACLs might be unwieldy and I also wonder about overhead of checking against the ACLs when some operation needs to take place. A practical example might be for library catalogue records, where each record gets its own graph. Records from a particular source, say a national library we might want to mark as read-only for unprivileged users, whereas user created information like reading lists and bibliographies should be writeable. A different use case from the one that spawned this thread since it doesn't involve completely separate table namespaces. Maybe it could be done with the clustering, where data held in one node has the modify bits removed? Else solve it at the application level... Cheers, -w -- William Waites <mailto:w...@styx.org> http://river.styx.org/ww/ <sip:w...@styx.org> F4B3 39BF E775 CF42 0BAB 3DF0 BE40 A6DF B06F FD45
