W3C home > Mailing lists > Public > www-svg@w3.org > November 2004

Re: Reconsider SVG 1.2

From: Jon Ferraiolo <jon.ferraiolo@adobe.com>
Date: Wed, 17 Nov 2004 20:22:21 -0800
To: Jim Ley <jim@jibbering.com>, www-svg@w3.org
Message-id: <>

As always, you make very good points.

Regarding your comments on SVG's specificity rules and about the issue of 
user agent stylesheets within a presentation grammar focused on final-form: 
Despite your good points, it is unlikely that SVG's use of CSS, SVG's 
specificity rules, and SVG's ability to deploy user agent stylesheets will 
get modified in SVG 1.2 whether or not you are correct. All of the 
decisions were made in the SVG 1.0 timeframe after much (sometimes very 
heated) debate, and the Web Accessibility team was involved in the process. 
We didn't touch any of those features with SVG 1.2, so almost certainly the 
SVG working group would say that these issues were closed in Sept. 2001 
when SVG 1.0 became a Recommendation and that it doesn't see a reason to 
reopen these heated discussions again.


At 07:28 AM 11/17/2004, Jim Ley wrote:

>"Philippe Lhoste" <PhiLho@GMX.net> wrote in message
> > That's why I am surprised by your statement that CSS is rarely used in
> > real world documents.
>The reasons it's rarely used are many fold, here's a few that I know of.
>    It's optional, many user agents do not support it, so if you're authoring
>for widest possible support including mobile, you don't use it - even if
>you're not targetting any player but ASV now, you don't cut yourself from
>the other UA's unless you have to.
>    Authoring tools avoid it, because it's not available everywhere, and it's
>more complicated to author ( you need to make sure that your specificity of
>your selectors is accurate, if you do that the easy way your CSS is no
>smaller as you don't get any re-use.)
>    It complicates scripting - because CSS properties are always higher
>specificity than attribute properties, it means you have to do
>setAttributeNS(null,'style','construct a string of all style
>operties')   -  this is both slower and more complicated than just setting
>the stroke-width attribute.
>My biggest complaint though, is the problem with the user stylesheet,
>because in CSS the user stylesheet rules supreme, the user stylesheet can
>render your images inaccessible, this I think shows clearly why CSS should
>not be used to style rendered semantics, and why the specificity of SVG
>attribute properties is wrong.  Hopefully the WG will address this issue of
>mine I've had since 1.1 as part of the 1.2 LC, as it obviously all applies
>to 1.2 as none of this has changed.
Received on Thursday, 18 November 2004 05:18:26 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 8 March 2017 09:47:01 UTC