W3C home > Mailing lists > Public > www-style@w3.org > September 2005

Re: [css3-text] Text shadow issues

From: Ian Hickson <ian@hixie.ch>
Date: Wed, 21 Sep 2005 18:07:19 +0000 (UTC)
To: Christian Biesinger <cbiesinger@web.de>
Cc: www-style@w3.org
Message-ID: <Pine.LNX.4.62.0509211744460.13412@dhalsim.dreamhost.com>

On Wed, 21 Sep 2005, Christian Biesinger wrote:
> I'm currently implementing text-shadow support. I have a few questions 
> about the specification text at 
> http://www.w3.org/TR/2003/CR-css3-text-20030514/#text-shadows (and the 
> pretty similar CSS2 wording, but given that this property was removed in 
> CSS 2.1, it will probably not be clarified in a further CSS 2 revision).

Here's the last unofficial consensus of the working group on this matter:

'text-shadow' is inherited.

Shadows should be painted for a glyph only as per the glyph's element's 
'text-shadow' property.

Shadows should be painted for a text decoration (underline, overline, etc) 
only according to the 'text-shadow' property of the element that has a 
non-none 'text-decoration' value.

Shadows should be painted immediately below the 'underline' text- 
decoration in the stacking order described in CSS2.1 appendix E. 
Specifically, insert "0. text-shadow of element's glyphs and decorations" 
as item 0 of subitem 4 of subitem 1 of subitem 6 in section E.2. The first 
shadow specified should be the lowest shadow painted, the last shadow 
specified should be the one painted "on top" of the others.

If a shadow's <color> is not specified, the UA may use whatever colour it 
believes is most appropriate. It is not required to use the value of the 
'color' property to paint the colour. Some implementations have found that 
for a common situation consisting of black text on a white background, the 
most aesthetically pleasing shadow is obtained from a dark gray.

The 'text-shadow' property should be honoured when set on the ::selection 

Hopefully the above will make it into an errata for text-shadow in due 
course. The CSS3 Text module currently in CR is scheduled be turned back 
to WD for serious maintenance work, including probably having it split 
into two specs (one for vertical text, one for the other text stuff). As I 
understand it we're waiting for the new spec(s) to be ready to republish 
before moving forward with this, at which point hopefully the exact rules 
for 'text-shadow' will be made clearer.

This reply was based on unpublished discussions in the CSSWG, primarily 
based upon these discussions and e-mails (member-only links):


It is believed to be what Safari has implemented. (I'm assuming the WG 
didn't discuss this at yesterday's teleconference, which I was unable to 
attend, and for which I have not yet the opportunity to read the minutes. 
If it was indeed discussed, then the WG's consensus on this matter may now 
be different than what I have described above.)

Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'
Received on Wednesday, 21 September 2005 18:07:39 UTC

This archive was generated by hypermail 2.3.1 : Monday, 2 May 2016 14:27:20 UTC