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

Re: Stand-in color before images load

From: Emrah BASKAYA <emrahbaskaya@hesido.com>
Date: Sat, 02 Apr 2005 16:02:20 +0300
To: "www-style@w3.org" <www-style@w3.org>
Message-ID: <opsolvt6gc8nstxa@lomarnona>

> Lachlan Hunt wrote:
>>>> I assume h1:alt would still match the H1 element if image loading
>>>> was somehow prevented.
>>  It wouldn't be necessary, doesn't this h1[src] { content:
>> attr(src,url); } already fallback to the normal content when that's
>> not available?
> Sure, but can you style the H1 element in a different way then? (Perhas  
> with that other proposal, but that was not discussed here.)

Exactly, that's a good point, what about background-attachment, repeat-x,
strech, etc etc.

On Wed, 30 Mar 2005 14:56:46 +0200, Laurens Holst
<lholst@students.cs.uu.nl> wrote:

> p { background: url("marble.png"), url("marble.gif"), transparent;  }
> Using a comma for alternates is used elsewhere in the spec as well (e.g.  
> content:), and there would be no need to introduce a new property.  
> What’s more, existing UA’s won’t recognise the syntax in this rule and  
> skip it, allowing you to provide another background rule for those with  
> perhaps a non-transparant background image as a fallback solution.
> Of the three, ‘transparent’ is the last alternate, but UAs can already  
> temporarily render it while one of the previous alternates is being  
> loaded.
> I think this would be a pretty nice solution, better than using :alt.
> ~Grauw

Using the exact same property as in CSS2 for the job is not feasible as it
might be really awfully interpreted by the old user agents.

I have checked this earlier, and current IE understands nothing with a
line with commas, always uses the last image in the line or the color if
declared. Opera interprets the last meaningful setting with or without

So in the end, we do need one or two extra properties (I preferer this for
the sake of simplicity), or maybe a pseudo-class.

I don't see the reason of complicating the markup anymore with selectors
and such. The probability of a selector or pseudoclass being implemented
in a user agent is much less then an actual property like
background-standincolor which does one thing and one thing only.

However, as much as some will hate me for this, I'd really like to have a
null pseudo-class that targets css3 capable agents only, e.g. p:css3 {} ,
which will work as nothing to interpreted by css3 agents. Why? We are not
living in an ideal world and before we have to use inefficient and hard to
understand 'hacks' that targets this and that, we should be able to
properly target css3 for reason we all know. (but hate to admit)


Using M2, Opera's revolutionary e-mail client: http://www.opera.com/m2/
Received on Saturday, 2 April 2005 13:27:07 UTC

This archive was generated by hypermail 2.3.1 : Monday, 2 May 2016 14:27:17 UTC