21558 2013-04-02 23:53:25 +0000 dependency wrong for "format-date-en152" and "format-dateTime-en152" 2013-05-01 09:17:38 +0000 1 1 1 Unclassified XPath / XQuery / XSLT XQuery 3 & XPath 3 Test Suite Working drafts PC Linux RESOLVED INVALID P2 normal --- 1 spungi mike mike public-qt-comments oldest_to_newest 85464 0 spungi 2013-04-02 23:53:25 +0000 <dependency type="calendar" value="CB" satisfied="false"/> should be <dependency type="calendar" value="CB" satisfied="true"/> for the following 2 test cases: "format-date-en152" from "fn-format-date" <test-case name="format-date-en152"> <environment> <param name="b" as="xs:date" select="xs:date('2006-03-01')"/> </environment> <dependency type="calendar" value="CB" satisfied="false"/> <test>format-date($b, '[M01]', 'en', 'CB', ())</test> <result> <assert-string-value>[Calendar: AD]03</assert-string-value> </result> </test-case> and "format-dateTime-en152" from "fn-format-dateTime" <test-case name="format-dateTime-en152"> <environment> <param name="b" as="xs:dateTime" select="xs:dateTime('2006-03-01T12:00:00')"/> </environment> <dependency type="calendar" value="CB" satisfied="false"/> <test>format-dateTime($b, '[M01]', 'en', 'CB', ())</test> <result> <assert-string-value>[Calendar: AD]03</assert-string-value> </result> </test-case> 87071 1 mike 2013-05-01 08:35:53 +0000 I don't agree. These tests demonstrate that if the implementation does not support calendar CB, it is expected to fall back to some other calendar such as AD. That is, the test is explicitly designed to be run by processors that do NOT support calendar="CB", which is what the metadata says. One could argue for alternative results (there's nothing in the spec that requires the fallback calendar to be "AD" or the word "Calendar" to be in English) but we can add alternatives, or additional dependencies, when we find products that make a different choice. 87074 2 mike 2013-05-01 09:17:38 +0000 Closing this as INVALID; if you disagree, feel free to re-open.