W3C home > Mailing lists > Public > www-validator-css@w3.org > February 2006

RE: CSS Validation warning: You have no background-color with your color;

From: Peter Normann <peter@normann.com>
Date: Thu, 23 Feb 2006 09:48:56 +0100
To: "'CSS validator list'" <www-validator-css@w3.org>
Message-ID: <074e01c63855$fb732200$062b96d5@normann.com>

Jukka K. Korpela wrote: 
> My concern was that it was not factually accurate. There
> are apparently other concerns, too.

Well, as long as it doesn't produce an error, accuracy must be measured
against what you actually are trying to accomplish.

> In almost all cases, the warning about potential problems
> when not specifying color and background together
> indicates lack of sufficient care. Almost all authors who
> complain about the warning _have_ actually failed to
> understand the problem, even after reading the
> explanation.  

I would agree entirely given the case that the warning was due to no
background being defined at all. But the warning is being produced even
though I specify a transparent background-color, thus telling the browser to
render the element with the background of the underlying element. It seems
to me a little far fetched that this is poor design by definition and/or
poor understanding of the issues of any correlation with an unknown

I concur that the design *might* be flawed if the author did not consider
the background-color of the underlying object could be something unexpected,
but in the real world you would not specify colors with really low contrast
to the background of the underlying background element when using

If the user agent throws some unknown stylesheet into the equation that skew
the contrast between the elements by deciding that one type of element
should have a black background and other types of elements should have white
backgrounds, thus silently ignoring any correlation between elements using
*specified* transparencies and colors, I plead guilty on all accounts. Maybe
I must admit that I do not really understand the real issues of the problem,
since my view of the situation is that the user in these cases should be
held liable for any deviation of the 'none or all' approach, as far as color
scheming is concerned. If they overrule my transparency background setting,
they should also overrule the color.

If using transparency is poor design per se, I move that the parameter
'transparent' should be removed from the CSS recommendation altogether and
taken outside and shot, since it obviously puts the internet at risk of
becoming unreadable.

> The continued complaints about the warnings indicate that
> they actually reveal design mistakes.

The world would be better of with no stylesheets or images on web pages at
all, come to think of it. No funny layout, no hazzles, just plain, hard
information for everybody to digest. Really. Hey, why did we even abandon
the good old monochrome high persistence phosphor monitors in the first
place? I kinda liked the cool after glow effect.

Don't get me wrong. It is not that I am being ignorant in regard to
accessibility. In fact I strongly believe that developers and web designers
should adhere to any recommandations in WAI or other resources with a take
on how to present information effectively to masses given uncertain

While I would like to see the background-color warning eliminated when the
transparency parameter is specified, I'ld love a "design validator" that
could offer suggestions to your styling to increase the likelyhood of john
doe getting the stuff presented on his screen.
Received on Thursday, 23 February 2006 08:55:43 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 21:40:41 UTC