- From: Lee Feigenbaum <lee@thefigtrees.net>
- Date: Tue, 06 Apr 2010 09:41:08 -0400
- To: SPARQL Working Group <public-rdf-dawg@w3.org>
Options that I see: O1) Include MINUS and NOT EXISTS as graph pattern clauses; include NOT EXISTS and EXISTS as FILTERs. => LeeF claims this is confusing pedagogically O2) Include MINUS as a graph pattern clause; include NOT EXISTS and EXISTS as FILTERs (original F2F3 decision) => AndyS claims that not having an "in-place" NOT EXISTS operation does not serve as an adequate replacement for the current OPTIONAL/!BOUND indiom. O3) Include MINUS graph operation only. O4) Include NOT EXISTS graph operation & NOT EXISTS/EXISTS FILTER only. => O3 and O4 fail to bridge the divide between people who view this as removing bindings versus filtering based on existence. O5) Include a NOT EXISTS graph operation. Define it in such a way that it covers the cases that have the same results for both approaches. Explicitly note that implementers may choose which style to implement for the corner cases. (I think this was Sandro's suggestion at F2F3.) => We're not sure how to specify this, I don't think. O6) Do not include anything for negation. => This doesn't seem to fly with our charter. I'll go through these on today's call, but in the absence of consensus we may simply end up whittling this list down based on what people's favorite options are, and then putting it to a group vote. Lee
Received on Tuesday, 6 April 2010 13:41:50 UTC