- From: Gregory Williams <greg@evilfunhouse.com>
- Date: Mon, 31 Oct 2016 08:20:58 -0700
- To: Andy Seaborne <andy@apache.org>
- Cc: public-sparql-exists@w3.org
> On Oct 31, 2016, at 7:51 AM, Andy Seaborne <andy@apache.org> wrote: > > On 30/10/16 02:07, Peter F. Patel-Schneider wrote: >> >> The how is the less important difference. Proposal A pushes the bindings to >> the "beginnings" of group graph patterns via the Initial(t) where they >> affect each construct on the group graph pattern. Proposal B pushes the >> bindings to each BGP. This ends up in effect pushing the bindings to the >> beginning of each group graph pattern because there is always a BGP (maybe >> empty) at the beginning of the translation of each group graph pattern > > It's a bit confusing to talk about group graph patterns and proposal B because group graph pattern do not exist in the algebra. > > Evaluation, being bottom up, always evaluates from BGPs upwards. BGPs are the point at which patterns access the data - "GRAPH ?g" is the other place. Also property paths. Of particular interest, I think, are zero-length paths that match graph nodes directly. .greg
Received on Monday, 31 October 2016 15:21:30 UTC