Re: [css syntax] CSS Syntax and Semantic Classes

CSS Working Group,




By expansions, I meant expansion from one XML tree to another, where often the web component is more succinct than the expanded, resultant, XML tree, expansion as per macro expansion (see also: XML macros http://www.w3.org/community/argumentation/wiki/XML_Macros).




I agree that CSS has convenient syntactic aspects and features, classes being one of them, conveniences resulting from group processes, brainstorming, in this mailing list.  I happen to like XHTML better than HTML (http://en.wikipedia.org/wiki/XHTML#Relationship_to_HTML, “The exact parsing of HTML in practice has been undefined until recently; see the HTML5 specification”, http://wiki.whatwg.org/wiki/HTML_vs._XHTML); in my opinion, HTML5 should have been and an HTML6 should be XML-based, XHTML, simplifying the Open Web Platform, Document Object Model and JavaScript development (https://developer.mozilla.org/en-US/docs/Writing_JavaScript_for_XHTML).




Kind regards,


Adam Sobieski




















































































































From: jackalmage@gmail.com
Sent: ‎Thursday‎, ‎February‎ ‎6‎, ‎2014 ‎2‎:‎22‎ ‎PM
To: Adam Sobieski
Cc: www-style@w3.org





On Thu, Feb 6, 2014 at 8:08 AM, Adam <adamsobieski@hotmail.com> wrote:
> CSS Working Group,
>
> The topics of semantic styling become more interesting with expansions of
> XML, e.g. web components.  Web components and styling web components are
> relatively new topics.  Web component authors, e.g. CMS widgets, might want
> to make use of the expressiveness of @xhtml:role, @rdf:type, @rdfa:typeof or
> @epub:type and possibly with semantic aspects of @class.  Semantic styling
> could also be of use for such scenarios.

Web Components aren't expansions of XML.  They're additions to HTML,
which can be written in XML via the XHTML variant syntax.

You can already use role='' and the RDFA stuff in your HTML document.
epub:type='' was a silly mistake that I think EPub is backpedalling
from now - they got infected with the XML namespaces meme and no one
caught them in time.

None of this has any relevance to your suggestion.  Classes are an
author-extensibility mechanism.  When standards have leaned on
classes, such as Microformats, it's been sufficient to just use short
prefixes and context.  Nobody has ever presented use-cases that needed
more disambiguation than that.

Plus, ultimately, classes are merely a convenient syntax for tagging
an element with an identifier, which is something you can do just as
easily with a binary attribute.  (The syntax isn't any more onerous,
either - <div foobar> and <div class="foobar"> are similar, as is
.foobar vs [foobar].)  Attributes already support namespaces, so that
provides the ability you're asking for already.  (You can't *write*
namespaced attributes in the HTML syntax, but you can produce them
with DOM methods, or you can just write in the XHTML syntax.)

~TJ