- From: Gérard Talbot <www-style@gtalbot.org>
- Date: Sat, 05 Nov 2016 21:23:11 -0400
- To: Felix Miata <mrmazda@earthlink.net>
- Cc: W3C www-style mailing list <www-style@w3.org>
Le 2016-11-05 20:01, Gérard Talbot a écrit : > Le 2016-11-05 19:03, Felix Miata a écrit : >> Gérard Talbot composed on 2016-11-05 17:47 (UTC-0400): >> >>> An inline style is (should be!) for an unique element. >>> Therefore, an inline style should be more specific than a class. >> >> That only works if the only styles are author, or if users aren't >> supposed to have ultimate control. > > Yes. If media (origin) is the same and if importance (user, author, > user agent with !important or not) is the same, then inline style is > more specific than a class attribute. > > https://www.w3.org/TR/CSS21/cascade.html#cascading-order > >> AFAICT, inline styles are >> hopelessly immune to user overrides, > > Hm... no. An author inline style can be (will be and should be) > overriden by an user !important class attribute. Untangling > specificity only comes after untangling importance. And I believe the > CSS 2.1 test suite has not tested this... can not find it in > > http://test.csswg.org/suites/css2.1/nightly-unstable/html4/chapter-6.htm#s6.4 > > [ > I think we need to test this... Something like: > > <body> > <p id="cascade">PREREQUISITE: The <a > href="support/cascade.css">"cascade.css"</a> file is enabled as the > user agent's user style sheet.</p> > <p>Test passes if "Text sample" is green.</p> > <h1 class="cascadegreenimportant" style="color: red;">Text > sample</h1> > </body> > > where the user style sheet is: > > http://test.csswg.org/suites/css2.1/nightly-unstable/html4/support/cascade.css > > ] Felix, I just tested this with Stylish 2.0.7 for Firefox https://addons.mozilla.org/en-US/firefox/addon/stylish/ and with userScriptCSS 1.4 for Chrome https://chrome.google.com/webstore/detail/userscriptcss/pdfbjinabdohnegjnbfgdgohlhegamnm on this test: http://www.gtalbot.org/BrowserBugsSection/css21testsuite/cascade-013.xht and it works. Firefox 49.0.2 and Chrome 54.0.2840.90 pass this test. Gérard
Received on Sunday, 6 November 2016 01:23:49 UTC