- From: Jukka K. Korpela <jkorpela@cs.tut.fi>
- Date: Mon, 10 Feb 2014 22:00:35 +0200
- To: www-validator@w3.org
- CC: "master.skywalker.88@gmail.com >> Andry Rendy" <master.skywalker.88@gmail.com>
2014-02-10 20:40, Andry Rendy wrote: > To the attention of validator.w3.org/nu <http://validator.w3.org/nu> staff. Hi Andry, I’m not affiliated with W3C in any way, just reading this list and trying to make some contributions. > But the specification states, as content model for > <iframe> in XML (and therefore in XHTML) that this element must be > empty. So it had to flag both cases as the same error, while the > requirements given in prose should be valid only in HTML (no test of > this case anyway). I can confirm that both validator.w3.org/nu and validator.w3.org and www.validator.nu pass an XHTML document that starts with <!DOCTYPE html> (triggering HTML5 validation in XHTML mode) containing <iframe src="foo">Hello world</iframe> I tested this using file upload with a filename ending with .xhtml, and the syntax check was thus performed by XHTML rules. Since HTML5 CR clearly says “The iframe element must be empty in XML documents”, this seems to be a bug in the validators. They apply the HTML rules even when for XHTML documents. Yucca P.S. The iframe element has no content model in XML—XML as such defines no content models. Its allowed content as per HTML5 is defined in the prose of HTML5 drafts (not in a DTD), so the prose cited in the question, and quoted above, is the only ground for deciding what is allowed in the content.
Received on Monday, 10 February 2014 20:01:06 UTC