A posting from Henry S. Thompson announces a "first public release of the W3C XML Schema Test Collection, made possible by a substantial contribution of tests from Microsoft. Both positive and negative expected outcomes are tested with respect to a range of core XML Schema features. [The tests are presented] in a standard form which tabulates (without ratifying) the test materials, together with a brief description, and the outcomes for each one expected by the contributor. The document also includes the first of what the W3C team hopes will be many outcome tabulations for a publically available XML Schema processor... the column labelled 'Expected' means the outcome expected by the contributor [not necessarily what's expected by the W3C WG]. For the test file(s) present which has/have extension .xsd, its/their conformance to the XML Schema REC's definition of valid XML representations of XML Schemas is what is at issue. When a test file with extension .xml is present as well, its schema-validity is at issue as well." Thompson reports that the W3C team already has in hand an additional contribution of tests from NIST; these will be added soon to augment the 100+ tests from Microsoft. Contributions from other sources are strongly encouraged. The test materials are available for download from the W3C web site as a single package, distributed under the W3C Document License.

"The W3C XML Schema Test Collection work aims at coordinating test suites for W3C XML Schema processors created by different developers. Here are its main objectives: (1) Integrate existing tests for W3C XML Schema processors in a common environment so they can be accessed publicly and shared among developers; (2) Establish a standard approach to test material IPR which meets the needs of both contributors and users; (3) Collect and develop tools to automate the execution and presentation of the test suites; (4) Offer a standard description of tests related to W3C XML Schema processors; (5) Test descriptions should be understandable by a developer without the need to actually view the test file(s) themselves; (6) Offer a standard presentation of test results; (7) Design additional tests and add/regularise descriptions of the existing tests; (8) In due course, provide an XSLT-based approach to comparing XML representations of the post schema-validation infoset as produced by different processors; we will shortly announce the availability of XML Schemas for both the ordinary Infoset and the PSVInfoset..."