Yes, Lifepillar's way is the more orthodox approach, however I always preferred the path-based one because: 1) One seldom runs queries only based on the descendants map; there usually is an "AND -some other conditions-" involved; thus the ability to have one covering index of the condition comes in very handy 2) In ontologies you very often also require "structural" sorting - such as, the chapters in a book may need to be read like 1<1.1<1.1.A<1.1.B<1.2 etc; this is very likely to be needed at some point and only achievable using some modified form of the path field (with alpa-sortable numbers). This cannot be achieved with a normalized descendency map.
-- Sent from: http://sqlite.1065341.n5.nabble.com/ _______________________________________________ sqlite-users mailing list sqlite-users@mailinglists.sqlite.org http://mailinglists.sqlite.org/cgi-bin/mailman/listinfo/sqlite-users
