W3C home > Mailing lists > Public > www-style@w3.org > April 2012

Re: [css3-values] syntax identity of <url> (was: [css4-values] Templated URLs)

From: Brian Kardell <bkardell@gmail.com>
Date: Fri, 6 Apr 2012 06:25:56 -0400
Message-ID: <CADC=+jfXi86KNazaz1yX54dZHgJOLdkrcJuD-NZShYr9+qqQTw@mail.gmail.com>
To: "Kang-Hao (Kenny) Lu" <kennyluck@csail.mit.edu>
Cc: fantasai <fantasai.lists@inkedblade.net>, WWW Style <www-style@w3.org>
Just wanted to note that as I am watching all of these threads fly by I
felt some dissonance between a few of these caused by the example used
here...

On the one hand, we have Tab's proposal for pseduo-element combinator...
it makes sense that ::after really is something wholly different,
especially as it relates to shadow dom and the components work.   On the
other hand it appears that this example wants to treat them as the same.
Is there a disconnect or is the example wrong?
On Apr 5, 2012 10:43 PM, "Kang-Hao (Kenny) Lu" <kennyluck@csail.mit.edu>
wrote:

> (12/03/21 7:20), fantasai wrote:
> > I received a suggestion[1] from Maxime Thirouin [1] for templating URLs
> > with the following example. [2]
> > Forwarding here for archival / discussion.
> >
> > ~fantasai
> >
> > [1] https://twitter.com/MoOx/status/173093535275421697
> > [2] http://dabblet.com/gist/1753404
> >
> > [data-gravatar-uri]::after
> > {
> >     content: url(attr(data-gravatar-uri)); /* impossible */
> > }
> >
> > [snip other new proposals]
>
> When it says 'impossible', does this mean browsers currently don't
> support it or CSS3 V&U doesn't support it?
>
> This relates to this questionable statement in the spec
>
>  # It corresponds to the URI token in the grammar. [CSS21]
>
> Is this really true? It was suggested that FUNCTION STRING ')' makes a
> <url> too[1] and there's even a test case in the CSS 2.1 test suite (see
> the link).
>
> But if 'url(' <string> ')' makes a <url>, I would suggest we include
> this description (like how image() is defined) so that the above works.
> (Or explicitly disallow it.)
>
>
> == other irrelevant nitpickings in this section ==
>
> I don't know if this section should remind spec editors that <url> was
> <uri> in CSS 2.1. I notice that CSS3 Images is, for instance, using the
> wrong ones. But in any case, I support changing the type name to be in
> line of the function name.
>
>  # Furthermore it replaces the ‘<url>’ type in the
>  #‘background-image’ and ‘list-style-image’ definitions in CSS1
>  # and CSS2 and adds ‘<image>’ as an alternative to ‘<url>’ in the
>  # ‘content’ property's value. It is presumed that CSS
>  # specifications beyond CSS2.1 will use the ‘<image>’ notation in
>  # place of ‘<url>’ where 2D images are expected.
>
> (This was too nitty so I didn't get to report it.)
>
>  # Note that in some CSS syntactic contexts (as defined by that
>  # context), a URL can be represented as a <string> rather than by
>  # <URL>. An example of this is the ‘@import’ rule.
>
> s/<URL>/<url>/
>
>  # Parentheses, whitespace characters, single quotes (') and double
>  # quotes (") appearing in a URL must be escaped with a backslash so
>  # that the resulting value is a valid URL token, e.g.
>  #‘url(open\(parens)’, ‘url(close\)parens)’.
>
> s/URL token/URI token/
>
> and also that should link to the CSS 2.1 grammar instead of the
> <dfn>URL</dfn>
>
> [1] http://lists.w3.org/Archives/Public/www-style/2011Apr/0701
>
>
>
> Cheers,
> Kenny
>
>
Received on Friday, 6 April 2012 10:26:26 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 17:20:52 GMT