W3C home > Mailing lists > Public > www-html-editor@w3.org > January to March 2006

Re: [XHTML 2] 24.2 Position of param elements (PR#7753)

From: Shane McCarron <shane@aptest.com>
Date: Wed, 25 Jan 2006 06:12:37 -0600
Message-ID: <43D76B35.6010708@aptest.com>
To: Anne van Kesteren <annevk@opera.com>
CC: Lachlan Hunt <lachlan.hunt@lachy.id.au>, Shane McCarron <xhtml2-issues@hades.mn.aptest.com>, jim@jibbering.com, www-html-editor@w3.org, xhtml2-issues@aptest.com

Anne van Kesteren wrote:

> On Wed, 25 Jan 2006 07:03:16 +0100, Shane McCarron <shane@aptest.com>  
> wrote:
>> I need to apologize.  I did not mean that I was looking for text for  
>> this one small section of the document. I could think of many 
>> perverse  situations for every element (what should a browser do when 
>> it finds a  div inside of a span?  a div inside of a p?).
> Given that creating a DOM for XML is more or less well defined that is 
> not  really a problem. At least, not from a visual point of view.

No.  The problem, just like PCDATA before a param element in an object 
element, is that it is invalid and therefore nonsense.  I know the XML 
processor could build a DOM tree.  You can build a DOM tree for anything 
that is well formed. But it remains invalid, and therefore wrong.

>> The point I was trying and failing to make earlier is that it is  
>> impractical to attempt to address all of these.  Actually, I maintain 
>> it  is by definition impossible.
> In my opinion it is possible. It is quite possible to define clear  
> processing rules that are both easy to implement and easy to write. 
> For  example, you could define that attributes that do not match the 
> production  that is defined for the attribute value MUST be ignored. 
> So that, UAs can  not by any means interpretid <div role="bogus"> as 
> <div role="navigation">.

The recommendation has rules for this already.  See the conformance 
section.  Here is an excerpt:

    If a user agent encounters an attribute it does not recognize, it
    must ignore the entire attribute specification (i.e., the attribute
    and its value).

    If a user agent encounters an attribute value it doesn't recognize,
    it must use the default attribute value.

> It is already shown by for example CSS 2.1 that this is not impossible 
> to  do. SVG Tiny 1.2 will eventually (and does already for a bit) show 
> the  same. I'm not sure why you think it is impossible given that it 
> is already  being done and appreciated a lot by implementors.
> No clear error handling is exactly what brought the web at the point 
> it is  today.

Fascinating.  And here I always thought it was lack of conformance to 
standards.  So-called "quirks mode" in user agents is what permitted 
crappy web sites to appear to work, and therefore to have to be 
supported by later, good browsers.  You of all people should agree with 

Imagine the state of the world if C compilers allowed syntactically 
incorrect C to be turned into object files.  Don't do it!  Break the 
cycle.  Your children will thank you.

Shane P. McCarron                          Phone: +1 763 786-8160 x120
Managing Director                            Fax: +1 763 786-8180
ApTest Minnesota                            Inet: shane@aptest.com
Received on Wednesday, 25 January 2006 12:13:34 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:08:54 UTC