- From: Juan Sequeda <juanfederico@gmail.com>
- Date: Tue, 17 May 2011 23:52:21 -0500
- To: public-rdb2rdf-wg@w3.org
- Message-ID: <BANLkTim0TEvupW01jtLW9bb=QcEa98iv1w@mail.gmail.com>
Hi everybody I quickly asked on twitter the following: "If you have a NULL in your relational database, what would you expect to see in RDF? Nothing? Create a triple with a bnode?" Here are some answers to my quick poll: just ignore it? Or is that too simplistic? http://twitter.com/#!/gothwin/status/70537526074540032 I would expect to see nothing. OWA sufficiently covers the "unknown" aspect of NULL http://twitter.com/#!/ldodds/status/70551817926344704 null equates to no value for me, important in db land, but w/ open world assumptions we have "don't know", so no triple imho http://twitter.com/#!/webr3/status/70554640441294848 i vote for no statement. http://twitter.com/#!/bhuga/status/70555667479535616 Aren't nulls mostly there due to the rigidity of RDBs? I'd say do nothing. Add column to schema, omit all nulls. http://twitter.com/#!/pablomendes/status/70565162964365312 I'd expect to see nothing; that is, the absence of a triple. http://twitter.com/#!/bendiken/status/70565965724790784 Seems like an unofficial consensus is to completely ignore NULL values. Juan Sequeda +1-575-SEQ-UEDA www.juansequeda.com
Received on Wednesday, 18 May 2011 04:53:08 UTC