W3C home > Mailing lists > Public > www-style@w3.org > March 2004

Re: [CSS3-hyperlink]Some comments on WD

From: Sam Kearns <sam@hinterlands.com.au>
Date: Wed, 10 Mar 2004 10:17:36 +1100
Message-ID: <404E5090.3010107@hinterlands.com.au>
To: www-style@w3.org

The points mentioned here by Liorean support my earlier assertion (on 
which I recieved no feedback at all) that opening and control of new 
windows/tabs/rendering surfaces should not be part of CSS (nor any other 
markup) at all. Giving this control to page authors was a mistake that 
needs to be undone. These decisions should *only* be made by the user or 
the user agent in accordance with browser/OS capabilities and user 

liorean wrote:

> 1. The 'target-name' property specifies some property values that you 
> might want to override in a user stylesheet. However, there is 
> currently no way to match elements based on a property-value 
> combination in css.
> You may currently override on a selector-for-selector basis, but none 
> that affects only a specific value for the property, nor any that 
> matches all elements that have a certain property. Both of these would 
> be wanted for full user control.
> Examples:
>  - A user wants to change all links with 'target-name' set to 'root' 
> or 'parent' to instead use 'new'.
>  - A user may want to change all links with 'target-name' set to 
> 'modal' to 'new'.
>  - A user may want to change all external links (absolute URI?) to 
> have 'target-new' set to 'new', but have all internal links (relative 
> URI?) set to 'tab'.
> 2. The 'target-new' property specifies 'window', 'tab' and 'none', but 
> the distinction is not made clear, nor the behavior a user agent that 
> doesn't have one or the other should use. How about a browser that 
> uses a hierarchical list instead of tabs, or a browser where you have 
> tabs dissociated with the browser window, allowing a combination of 
> spatial orientation by windows, and tabs? And how do you handle tab 
> groupings?
> 3. The 'target-position' works like z-index, but for application 
> windows, as I understand it. However, isn't that very OS/GUI 
> dependent? How would a GUI that doesn't allow overlapping windows 
> handle them, or a GUI that only allows single windows, but provides a 
> tabbed application switching mechanism?
> 4. I agree with Anne van Kesteren, a mechanism for declaring generated 
> content to be a link would be a very nice addition to it.
Received on Tuesday, 9 March 2004 18:00:55 UTC

This archive was generated by hypermail 2.3.1 : Monday, 2 May 2016 14:27:12 UTC