- From: Kay, Michael <Michael.Kay@softwareag.com>
- Date: Tue, 10 Jun 2003 17:49:36 +0200
- To: public-qt-comments@w3.org
Software AG is concerned that the definition of a schema-aware processor, as defined in the conformance rules, should not be drawn too rigidly. The "import schema" mechanism is described as if a query can be compiled against an arbitrary schema located anywhere on the web. Software AG intends to take a much more constrained approach: all schemas that are used in a query must be preloaded into the schema repository of the target database, and the schemas that are available for use when compiling a query are precisely those that are registered in the database and that have been used for validating the documents that the query is run against. By default, our product will implicitly import all the schemas registered in the database against which a query is compiled. The "import schema" syntax can be accepted, for interoperability, but it is essentially redundant. We believe that this is a reasonable architecture that meets a wide range of user requirements, and that it should be considered conformant to the specifications. Apart from anything else, it solves the thorny problems of consistency between compile-time schemas and run-time schemas, which the working group is still wrestling with. Michael Kay Software AG
Received on Tuesday, 10 June 2003 11:49:50 UTC