W3C home > Mailing lists > Public > www-style@w3.org > October 2004

Re: Selector Decendant/Child Grouping Suggestion

From: Christian Roth <roth@visualclick.de>
Date: Tue, 5 Oct 2004 15:03:44 +0200
To: "www-style Mailing List" <www-style@w3.org>
Message-Id: <20041005140345.32187@mail.visualclick.de>

David Woolley wrote:

>As it doesn't add any new capability, this can be implemented by 
>the authoring tool, or a standalone pre-processor.

This is an interesting answer. If human writability and legibility is not
a main concern of CSS, but only new capabilities, then

- why are there still new shorthands introduced in CSS3 ('word-break',
'text-overflow', 'line-grid' etc.)? They're not adding new capabilities
and can all be generated by the authoring tool or a preprocessor.

- why is the parameter to :nth-child() not a string to ease parsing the
argument? The authoring tool can hide that fact from the user. A similar
issue holds true for the url() function.

- why is there no standardized XML representation of a CSS stylesheet,
eliminating  the need to define syntactic error handling? The authoring
tool can well hide the underlying representation (CSSOM and its
serialization, e.g.) and structural errors can no longer happen, resp.
are caught by XML validators or already the authoring tool.

If the human readability of the CSS syntax is not of highest importance,
then we should seriously consider leveraging existing technologies to
make parsing CSS easier: Make a cut for CSS3 and introduce an easily
parseable, consistent, clear and single (=normalized) syntax with
rigorous error handling. Note that this does NOT mean to invent a
completely new syntax, but it's mostly small changes that  make the
difference.

W3C's CSS validator still isn't a fully conforming CSS2 parser - isn't
that indicative of some issue here?

Regards, Christian.
Received on Tuesday, 5 October 2004 13:04:47 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 27 April 2009 13:54:34 GMT