Hi Ivan, Here is a dataset: http://www.mediafire.com/?drtwizwozq5
Here are sample queries to compare: prefix geo: <http://www.w3.org/2003/01/geo/wgs84_pos#> prefix geonames: <http://www.geonames.org/ontology#> SELECT DISTINCT ?cityName FROM <http://www.geonames.org#> WHERE { ?city ?predicate ?value . ?city geonames:featureClass ?featureClass . ?city geonames:name ?cityName . ?value bif:contains '"tomsk"' . FILTER (?featureClass = "http://www.geonames.org/ontology#P") } and prefix geo: <http://www.w3.org/2003/01/geo/wgs84_pos#> prefix geonames: <http://www.geonames.org/ontology#> SELECT DISTINCT ?cityName FROM <http://www.geonames.org#> WHERE { ?city ?predicate ?value . ?city geonames:featureClass "http://www.geonames.org/ontology#P" . ?city geonames:name ?cityName . ?value bif:contains '"tomsk"' . } Regards, Alexander 2010/4/24 Ivan Mikhailov <imikhai...@openlinksw.com> > Hello Alexander, > > What is the complete text of the query and what are the data in > question? Additionally, what is exact server version? The case is quite > strange because "by default" these two should result in identical > fragments of SQL. > > Best Regards, > > Ivan Mikhailov > OpenLink Software > http://virtuoso.openlinksw.com > > > On Fri, 2010-04-23 at 23:28 +0700, Alexander Sidorov wrote: > > I forgot to notice that first query doesn't work. > > > > 2010/4/23 Alexander Sidorov <alexi...@gmail.com> > > Hello! > > > > Should these queries: > > > > WHERE > > { > > ?s <http://www.geonames.org/ontology#featureClass> > > "http://www.geonames.org/ontology#P" . > > } > > > > WHERE > > { > > ?s > > <http://www.geonames.org/ontology#featureClass> ?featureClass . > > FILTER (?featureClass = "http://www.geonames.org/ontology#P") > > } > > > > be equivalent? > > > > Yes, I know featureClass predicate stores URIs but not > > literals at GeoNames dataset. But my custom GeoNames dataset > > first stored it as literals (it really doesn't change anything > > from semantic point of view as there is now URI > > <http://www.geonames.org/ontology#P>). Then for performance > > reasons I have reimplemented RDF generator so that > > featureClass stores URIs. But there is a chance this problem > > can be reproduced on more real usecases so I decided to report > > it. > > > > Regards, > > Alexander > > >