- From: Daniel Holbert <dholbert@mozilla.com>
- Date: Thu, 22 May 2014 09:16:45 -0700
- To: "Tab Atkins Jr." <jackalmage@gmail.com>
- CC: www-style <www-style@w3.org>
(Yup, I figured it was something like that.) So, the specifics of these test-bugs aside -- where should folks report bugs with test-modifications like this, so that they can be fixed? And where are the "live" versions of these tests tracked? Right now, the test being served by the live test suite doesn't seem to correspond to what's in Shepherd, and it also doesn't exist in the csswg "test" mercurial repo. (and hence, there's no way to view its change-history & find out when/why things broke, as far as I know.) ~Daniel On 05/22/2014 12:38 AM, Tab Atkins Jr. wrote: > On Thu, May 22, 2014 at 10:10 AM, Daniel Holbert <dholbert@mozilla.com> wrote: >> The problem is that there's a <!-- --> comment in the <style> block, >> which is interpreted as a comment in XHTML but apparently not in HTML, >> and it neuters some style in the converted-to-HTML version. > > Yeah, in HTML the contents of <style> are automatically CDATA and > passed through to the CSS parser literally. The CSS parser accepts > and ignores the top-level <!-- token, but the text following that > starts up a selector that continues up to and including the > .centerparent selector of the following rule. > > It's usually a good pattern to manually declare <style> as CDATA in > XHTML, which would have the same behavior, but as long as no > > combinators are used or any other troublesome characters in strings, > it's not necessary, so HTML comments work, since I think only the text > node children of the <style> block get sent through to the CSS parser. > > ~TJ >
Received on Thursday, 22 May 2014 16:17:12 UTC