W3C home > Mailing lists > Public > www-style@w3.org > January 2003

Re: XBL is (mostly) W3C redundant, and CSS is wrong W3C layer for semantic behavior *markup

From: Ian Hickson <ian@hixie.ch>
Date: Sun, 5 Jan 2003 23:42:15 +0000 (GMT)
To: Shelby Moore <shelby@coolpage.com>
Cc: Rijk van Geijtenbeek <rijk@iname.com>, "www-style@w3.org" <www-style@w3.org>
Message-ID: <Pine.LNX.4.21.0301052324280.4908-100000@dhalsim.dreamhost.com>

On Sun, 5 Jan 2003, Shelby Moore wrote:
> 
> However, if you instead used the CSS "display" property to change a
> <li> into a <table> element, then you would be creating
> non-conforming semantics.

CSS is unable to do this. Even XBL is unable to do this.

All CSS and XBL can do is change the _rendering_ of the element so
that it happens to use table-like layout.

In any case, there isn't any clause in the HTML specification that
says that rendering <li> elements with table layout is wrong. (I
checked.)


> This is why the "display" property is "not required by HTML
> conforming user agents"[1].

This is an uninformed lie. You have already been given the real reason
by two members of the working group, including the chair at the time:

   http://lists.w3.org/Archives/Public/www-style/2003Jan/0119.html

The clauses were added as "limp clauses", that is, in order to allow
implementors to claim conformance before having implemented the
properties for HTML elements. It was effectively a marketing driven
decision and had no relation to smenatic concerns.

The clause is in the process of being removed from CSS2.1.


> Yes I agree. Mixing "presentation" and "non-comforming semantic
> implementation" in inseparable ways will have some bad effects.

You have been given multiple ways in which CSS, according to your
logic, does this already. Yet I have yet to see any of these bad
effects.

I certainly don't see how XBL, or any other behaviour technology,
including HTCs or even transformation languages like XSLT and STTS,
could make this worse.


>> But I'm even more suspicious of efforts that promote sending
>> generic XML over the net for general purpose documents.
> 
> If I markup all the dates in my document using a custom tag
> <mydates>, it doesn't cause any harm.

This is a huge misconception of the XML groupies. It _does_ cause
harm. It means that implementations that do not apply your style hints
will totally fail to convey the document in any reasonable way.

You can actually see this now with documents that use <table> elements
for layout, or <div> elements for styling. Non-visual and non-CSS UAs
respectively don't have any chance of rendering the document usefully.

This is why iti is criticially important to have everyone involved or
even potentially involved in a transaction (which, or typical Web
servers, means everyone on the Web) agree on what the semantics of
elements are. Which is why the semantics cannot be affected by
implementations: the implementations come _after_ agreement has been
reached. (Because specs are implemented after being written.)

-- 
Ian Hickson                                      )\._.,--....,'``.    fL
"meow"                                          /,   _.. \   _\  ;`._ ,.
http://index.hixie.ch/                         `._.-(,_..'--(,_..'`-.;.'
Received on Sunday, 5 January 2003 18:42:16 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Monday, 27 April 2009 13:54:19 GMT