Re: 'role' should be property

----- Original Message ----- 
From: "Maciej Stachowiak" <mjs@apple.com>
To: <mark.birbeck@x-port.net>
Cc: <public-html@w3.org>
Sent: Wednesday, May 23, 2007 12:25 PM
Subject: Re: 'role' should be property


> 
> 
> On May 23, 2007, at 6:12 AM, Mark Birbeck wrote:
> 
>>
>> Hi Dmitri,
>>
>>> I am pretty sure the whole thing is a horrible idea. Not only the
>>> separation of concerns is compromised, but think about it: in  
>>> order to
>>> understand semantics of an HTML document, you'll also need to parse
>>> the stylesheet. Now you're doubling the effort for non-browser user
>>> agents. Semantics should stay in the markup and certainly not require
>>> additional documents to parse in order to understand the meaning of
>>> the document.
>>
>> Mmm...this doesn't sound like a discussion about software development
>> to me; something is a "horrible idea"? That means nothing. Some
>> software has to parse two things instead of one? Big deal...computers
>> spend most of their time doing nothing anyway. (Unless you count
>> sitting in a loop waiting for something to do as doing something.)
>> "Semantics should stay in the mark-up"? Which book of rules was that
>> handed down from? Semantics *about* documents, that are not themselves
>> contained within documents make up an enormous part of the web.
> 
> The problem with the idea is that it goes against the Architecture of  
> the World Wide Web. In particular it seems to go against 4.3.  
> Separation of Content, Presentation, and Interaction <http:// 
> www.w3.org/TR/webarch/#pci>. "Some data formats are designed to  
> describe presentation... these data formats should only address  
> presentation issues."
> 

Separation of Content, Presentation, and Interaction is a good thing.
Another good thing is single tool or technology that allows to do
Integration of Content, Presentation, and Interaction.  

As more that Integration tool is flexible as more it allows to isolate 
content from presentation or interaction styles so to do that Separation.

About the role: it should be role DOM attribute and role CSS atribute
that allowed to override value of the DOM attribute. Some element roles 
are known for the author and some only makes sense for 
particular domains (media).

IAccessible::accRole() [1] is a good example of the case.
Currently is just impossible for the web author willing to deliver 
rich content to make it yet accessible. 

Andrew Fedoniouk.
http://terrainformatica.com

[1] http://msdn2.microsoft.com/en-us/library/ms696152.aspx


> 
>> I'm not saying my idea is 'right' and there should be no debate. What
>> I'm saying is that if people want to debate issues like this, it is
>> important to get beyond simple assertions, that usually only reflect
>> what their proponent is familiar and comfortable with.
> 
> I'm not going to claim the TAG is automatically right about  
> everything they have ever said. But I think this particular finding  
> is widely shared, at least to the extent that styling languages  
> should be considered primarily presentational. It may be that  
> discarding this principle would have some value, but it would  
> certainly be a big deal, not something at the level of a trivial  
> syntactic change.
> 
> Regards,
> Maciej
> 
>

Received on Wednesday, 23 May 2007 21:13:00 UTC