W3C home > Mailing lists > Public > public-html@w3.org > July 2007

Re: What happens if ... we drop @style (was: Re: Wiki page for style attribute)

From: Robert Burns <rob@robburns.com>
Date: Wed, 4 Jul 2007 16:11:06 -0500
Message-Id: <22AF8F24-8A6F-4675-9C63-7D6634B81E9D@robburns.com>
Cc: public-html@w3.org, www-style@w3.org
To: Jens Meiert <jens.meiert@erde3.com>


On Jul 4, 2007, at 11:23 AM, Jens Meiert wrote:

>
> I'm sorry (only due the fact that I just love maintaining sites  
> that truly separate structure from presentation), but I judge the  
> @style discussion as becoming far from reality.
>
> So here's some "what happens if" concerning the "style" attribute.
>
> [...]
>

One thing to keep in mind is that anything dropped from HTML5 is  
still part of the UA conformance criteria. This means that @style  
will be there as a bridge.

On the other hand, I think that dropping @style is just one step  
further than I would take the attempt to be pure. Drop <font> by all  
means. Don't even allow it for legacy content editors. Obviously  
<font> will still be there for those editors, but we shouldn't  
endorse it. However, I think @style is something that we should  
provide for certain left-over cases. When separating presentation  
from semantics is just more cumbersome than is necessary. Separating  
presentation and semantics is most desirable when styling semantic  
and reusable markup. When its clear that style will only apply to  
this particular element, just this one time, its just a concern for  
theoretical purity that want so require an author to add an @id and  
then add a corresponding selector to a separate stylesheet.

On the other hand, keeping <font> around encourages authors to use an  
element that has no semantic effect and has a long practice of  
avoiding semantics. Often this is compared to <span>. and <div>,  
However, authors are already accustomed to adding semantic @class  
information to those elements. I think if we did one of these  
unscientific surveys we'd find many more <font> elements providing no  
semantic information than we found <div> and <span> elements without  
any semantic information (e.g., class="something-meaningful").

Again, these arguments are separate from what you're saying above. I  
think @style can be dropped from HTML5 and still serve as that bridge  
you spoke of as a part of the UA implementation conformance criteria.  
Dropping it merely says we think should not be considered a part of  
best practice authoring.

Take care,
Rob
Received on Wednesday, 4 July 2007 21:11:23 UTC

This archive was generated by hypermail 2.3.1 : Monday, 29 September 2014 09:38:46 UTC