- From: Leora Morgenstern <leora@us.ibm.com>
- Date: Tue, 11 Nov 2008 12:07:30 -0500
- To: public-rif-wg@w3.org
- Message-ID: <OFE0965233.2EF5F427-ON852574FE.005DD33E-852574FE.005E1105@us.ibm.com>
This email is intended to start a discussion about whether the current suite of test cases (available at http://www.w3.org/2005/rules/wiki/Category:Test_Case ) is sufficient for RIF's needs. I am attaching a spreadsheet that summarizes these test cases. (This is all taken directly from the wiki; the spreadsheet is just meant to make it easier to see at a glance what the test cases address.) Column C gives the main point of these test cases. Column E gives the reference to part of the spec in the particular RIF document (e.g., BLD, SWC, DTB) which the test addresses. Some parts of these specs seem to have more coverage than others. --- For example, only some of the built-ins in DTB are covered. Should there be test cases for all built-ins? --- Another example: There are no tests that point to the SWC section on entailment, although there are several examples pointing to the closely-related section on Common RIF-RDF Interpretations. Is this sufficient? --- Another example: Most UCR cases are not represented in the test suite. This is because most of the test cases can't be represented in BLD. Nevertheless, it raises the question: should there be other practical use cases which demonstrate how RIF will be used, and should there be test cases corresponding to these new use cases? The general question is: Does the current suite of test cases adequately cover the RIF document specs? Comments on this issue are welcome. Best regards, Leora
Attachments
- application/vnd.ms-excel attachment: RIFtest_case_features.xls
Received on Tuesday, 11 November 2008 17:08:18 UTC