[ https://issues.jboss.org/browse/ARTIF-722?page=com.atlassian.jira.plugin.... ] Brett Meyer updated ARTIF-722: ------------------------------ Description: Now that queries return a summary DTO, if a query only contains full-text search and columns indexed by Lucene, Hibernate Search can be solely used. There is no need to also execute the SQL query. For example: /s-ramp/xsd/XsdDocument[xp2:matches(., 'FizzBuzz')] Since 'model' and 'type' are both indexed, this could be accomplished entirely with the Lucene query, skipping SQL. This idea implies that all properties included in ArtifactSummary would need to be indexed as well. SQL would be needed only when custom properties, ontologies, or relationships are required. was: Now that queries return a summary DTO, if a query only contains full-text search and columns indexed by Lucene, Hibernate Search can be solely used. There is no need to also execute the SQL query. For example: /s-ramp/xsd/XsdDocument[xp2:matches(., 'FizzBuzz')] Since 'model' and 'type' are both indexed, this could be accomplished entirely with the Lucene query, skipping SQL. This idea could be expanded to include most built-in properties. Name, description, content hash, etc. could all be indexed as well. SQL would be needed only when custom properties, ontologies, or relationships are required. -- This message was sent by Atlassian JIRA (v6.3.15#6346)