W3C home > Mailing lists > Public > www-style@w3.org > March 2011

Re: [css3-background] Default shadow color

From: Brad Kemper <brad.kemper@gmail.com>
Date: Fri, 4 Mar 2011 09:45:23 -0800
Cc: Tab Atkins Jr. <jackalmage@gmail.com>, fantasai <fantasai.lists@inkedblade.net>, "www-style@w3.org" <www-style@w3.org>
Message-Id: <2304B47F-9F97-4769-9632-FF2A3A452515@gmail.com>
To: Sylvain Galineau <sylvaing@microsoft.com>

On Mar 4, 2011, at 9:08 AM, Sylvain Galineau wrote:

> [Brad Kemper:]
>> I'd rather leave it undefined in the hope that perhaps some specialized UA
>> could do the right thing (while the big guns Web browsers maintain status
>> quo if it is so much to ask to change), than to mandate a useless default
>> that can render text unreadable. And I'd expect consistency of the default
>> between box- and text-shadow.
> Keeping something undefined to make it more useful sounds like a dubious 
> trade-off for standard spec. 
> There may be good reasons to keep this undefined; not documenting a behavior 
> that all implementations actually agree on in the *hope* that this gives some
> unspecified imaginary someone an extra incentive to break interop to do something 
> no current author or existing stylesheet will benefit from doesn't sound like an 
> actionable priority to me.

Changing the default to currentColor also is something no current author or existing stylesheet will benefit from, except in the case of the author style sheet hack for thin letter strokes. 

Perhaps making the color required would be more palatable, since authors always include it anyway? Then you'd be documenting the behavior of all authors instead of all implementors.

> Any authoring environment that assists with writing CSS - from Adobe's product to 
> Daniel's BlueGriffon - is perfectly able to provide interesting smart shadow defaults 
> based on all the available information to them combined with any heuristics they 
> choose. Nothing prevents them from doing that that and I don't see how purposefully 
> avoiding implementation reality adds any value to that process.
> In the meantime, I'd like to test that the color component of shadows is optional. 
> That's a bit harder if transparent is a conformant default value. When everyone 
> already uses a perfectly testable behavior it also seems silly.

Everyone? Are you sure? We've tested Prince, et al?
Received on Friday, 4 March 2011 17:45:58 UTC

This archive was generated by hypermail 2.3.1 : Monday, 2 May 2016 14:38:44 UTC