tag:blogger.com,1999:blog-3322141.post112881107387504681..comments2023-10-24T23:22:27.416+10:00Comments on More News: JRDF 0.3.4.1 OutUnknownnoreply@blogger.comBlogger2125tag:blogger.com,1999:blog-3322141.post-1128833883713246052005-10-09T14:58:00.000+10:002005-10-09T14:58:00.000+10:00I'm just going to show how I misunderstand both th...I'm just going to show how I misunderstand both the RDF and Ruby (and other languages) here...<BR/><BR/>Getting duplicate blank nodes from the same statements from the same file "seems like a mistake" to me is a lot like "duck typing". It walks like a duck, it quacks like a duck - it is a duck.<BR/><BR/>I guess I'm just arguing that returning "infinity" to a count of statements on a graph with blank nodes does not seem quite right.<BR/><BR/>Ideally, it means that it isn't the default behaviour in APIs and query languages. <BR/><BR/>Unfortunately, this behaviour is also easiest thing to implement.Andrewhttps://www.blogger.com/profile/00667948202593884438noreply@blogger.comtag:blogger.com,1999:blog-3322141.post-1128820692412032402005-10-09T11:18:00.000+10:002005-10-09T11:18:00.000+10:00...in Kowari loading the sample FOAF files twice i...<I>...in Kowari loading the sample FOAF files twice into the same graph adds the blank nodes twice. While this seems like a mistake it is the fastest way to load a graph.</I><BR/><BR/>I'm glad you only said "seems like a mistake", because this behaviour is <B>required</B> by the RDF semantics. Note that it is entirely legal to map two (or more) blank nodes to the same name.<BR/><BR/>Of course, it is desirable to automatically find which blank nodes are the same as each other (owl:sameAs). In fact, I think it would be great if we could eliminate all redundant blank nodes. After all, the semantics allow for an infinite number of distinct blank nodes for every named node in the system... which would cause significant problems when stored in a real database.Paulahttps://www.blogger.com/profile/03653112583629043593noreply@blogger.com