14020 2011-09-04 01:27:53 +0000 HTML5+Microdata documents are not conforming HTML5 documents 2015-08-23 07:07:05 +0000 1 1 1 Unclassified HTML Checker General unspecified All All RESOLVED FIXED http://dev.w3.org/html5/spec/infrastructure#other-applicable-specifications P2 normal --- 1 xn--mlform-iua mike cpigat242 julian.reschke lrosenth public-html-admin www-validator-cvs oldest_to_newest 56260 0 xn--mlform-iua 2011-09-04 01:27:53 +0000 BACKGROUND: The W3C Validator checks for 'Conforming HTML5 documents" [0a]. Whereas Validator.nu checks for 'Conforming documents' per the Living Standard [0b]. ISSUE: The W3 validator's HTML5 conformance checker service currently does not discern between conforming HTML5 documents and conforming HTML5+Microdata documents - for both of them, the validator emits the same stamp: "This document was successfully checked as HTML5!" This contradicts with the outcome of the HTMLWG's ISSUE-140 [1][2], after which HTML5' Extensibility section (see especially 'other applicable specifications' [3] and onwards]) makes it clear that a document which implements an applicable spec with HTML5-different semantics, is not a conforming HTML5 document. Which is exactly the situation we have in the case of HTML5+Microdata: When Microdata syntax is applied, a group of HTML5-invalid and HTML5-semantic-free attributes are changed into HTML5+Microdata-valid and HTML5+Microdata-semantic attributes. It is not against the HTML5 spec to perform conformance checking of an HTML5+Microdata profile - on the contrary. But it is against HTML5 for the validator to pretend as if documents augmented with Microdata syntax have been "successfully checked as HTML5" - they have not. It would however be in line with the spec to state that they have been "successfully checked as HTML5+Microdata". The W3 Validator should be brought in line with the HTML5 spec ASAP. The borders of the spec - and its extensibility options, are two sides of the same coin. I suppose that Validator.w3.org as a minimum must offer pure HTML5 conformance checking. And thus that HTML5+Microdata checking would be an additional conformance checking service. Howeveer, this bug can formally be solved in 2 ways: 1. by starting to stamp HTML+Microdata documents as invalid. (This option also cover the possibility to start to offer separate conformance checking for HTML5 docs as well as HTML5+Microdata docs.) 2. by halting to offer HTML5 conforance checking - and only offer HTML5+Microdata checking The 1st option seems like the logical one. Given that there is an effort to join RDFa and Microdata [4], it might also - for the time being, be best to stamp Microdata as invalid instead of creating a new validation profile. [Which alludes to a third way to solve this bug: that microdata/RDFa becomes part of HTML5.] [0a] http://dev.w3.org/html5/spec/infrastructure#conforming-html5-documents [0b] http://www.whatwg.org/specs/web-apps/current-work/multipage/infrastructure.html#conforming-documents [1] http://lists.w3.org/Archives/Public/public-html/2011Mar/0574#start574 [2] http://www.w3.org/Bugs/Public/show_bug.cgi?id=9178 [3] http://dev.w3.org/html5/spec/infrastructure#other-applicable-specifications [4] http://www.w3.org/mid/4E04A795.5020609@arcanedomain.com 63885 1 mike 2012-02-08 12:48:42 +0000 The Nu Markup Validator default checker supports both Microdata and RDFa and also has an option for HTML5-only checking, for those who want to opt-out of Microdata and RDFa altogether.