Validatingresult strange

[Peter]

The stylesheet submitted for checking by W3C gives 2 warnings. no color
for background. But the backgroudcolor is
given. What can be wrong? This is the stylesheet:
http://lehmann.fateback.com/lehmann.css

--
http://lotto924.fateback.com

[Beauregard T. Shagnasty]

Peter wrote:

The stylesheet submitted for checking by W3C gives 2 warnings. no color
for background. But the backgroudcolor is
given. What can be wrong? This is the stylesheet:
http://lehmann.fateback.com/lehmann.css

You need to set *color* for those places that only set *background*.

body { color: black; }

#footer { color: transparent; }

--
-bts
-Warning: I brake for lawn deer

[Steve Pugh]

Peter wrote:

The stylesheet submitted for checking by W3C gives 2 warnings. no color
for background. But the backgroudcolor is
given. What can be wrong? This is the stylesheet:
http://lehmann.fateback.com/lehmann.css

The actual warnings are:

Line : 4 (Level : 1) You have no color with your background-color :
body
Line : 67 (Level : 1) You have no color with your background-color :
#footer

and the styles in question are:

body {
font-family: "Times New Roman", Times, serif;
width: 80%;
background: #FFC;
}

#footer {
background: #FFC;
border-top: 0.2em solid #CCE5FF;
clear: both;
width: 100%;
margin: 4% auto;
padding: 5%;
text-align: center;
font-size: 80%;
}

In both case you have set a background colour but not set a foreground
colour, hence the warning.

Steve

[Harlan Messinger]

Peter wrote:

The stylesheet submitted for checking by W3C gives 2 warnings. no color
for background. But the backgroudcolor is
given. What can be wrong? This is the stylesheet:
http://lehmann.fateback.com/lehmann.css

It doesn't say "no color for background". It says "You have no color
with your background-color : body" and "You have no color with your
background-color : #footer". These are accurate statements. You defined
background-color for both of these, and color for neither.

[Peter]

Harlan Messinger wrote:

Peter wrote:

The stylesheet submitted for checking by W3C gives 2 warnings. no
color for background. But the backgroudcolor is
given. What can be wrong? This is the stylesheet:
http://lehmann.fateback.com/lehmann.css

It doesn't say "no color for background". It says "You have no color
with your background-color : body" and "You have no color with your
background-color : #footer". These are accurate statements. You
defined background-color for both of these, and color for neither.

Thank you all for the quick response. It validates now!
Peter

--
http://lotto924.fateback.com

[Alan J. Flavell]

On Wed, 15 Feb 2006, Beauregard T. Shagnasty wrote:

Peter wrote:

The stylesheet submitted for checking by W3C gives 2 warnings. no color
for background. But the backgroudcolor is
given. What can be wrong? This is the stylesheet:
http://lehmann.fateback.com/lehmann.css

FAQ, see http://www.websitedev.de/css/validator-faq#color
You need to set *color* for those places that only set *background*.
It's advisable and recommended, indeed - for more-reliable cascading
with other stylesheets, particularly a user stylesheet or default
browser settings. And it's more resilient against some browser bugs.
body { color: black; }
I think you're saying to *add* that property setting to the existing
body { } specification, yes? Not to create two separate body { }
specifications...
#footer { color: transparent; }

That's not such a good idea. "transparent" runs much the same risk of
cascading badly with other stylesheets, as not specifying it at all
(in detail its cascading properties would be different, but it's
nevertheless the same principle).

It's best if you either specify *both* colours explicitly, or
*neither*, *at each specificity*. It isn't always feasible to do that
in every situation - especially when you have a background image on an
outer element which you want to shine through some inner element(s) -
but it's a good principle to keep to, as far as possible. This is a
*warning*, after all, rather than an actual error.

(See also the discussion under
http://www.w3.org/Bugs/Public/show_bug.cgi?id=2421 )

[Harlan Messinger]

Peter wrote:

Harlan Messinger wrote:

It doesn't say "no color for background". It says "You have no color
with your background-color : body" and "You have no color with your
background-color : #footer". These are accurate statements. You
defined background-color for both of these, and color for neither.

Thank you all for the quick response. It validates now!

You're welcome. I just wanted to point out that it validated before.
Those were warnings, not validation errors.

[Beauregard T. Shagnasty]

Alan J. Flavell wrote:

On Wed, 15 Feb 2006, Beauregard T. Shagnasty wrote:

Peter wrote:

The stylesheet submitted for checking by W3C gives 2 warnings. no color
for background. But the backgroudcolor is
given. What can be wrong? This is the stylesheet:
http://lehmann.fateback.com/lehmann.css
FAQ, see http://www.websitedev.de/css/validator-faq#color
Yes, I've seen it, but perhaps the OP hasn't.
You need to set *color* for those places that only set *background*.

It's advisable and recommended, indeed - for more-reliable cascading
with other stylesheets, particularly a user stylesheet or default
browser settings. And it's more resilient against some browser bugs.

body { color: black; }

I think you're saying to *add* that property setting to the existing
body { } specification, yes? Not to create two separate body { }
specifications...

Correctamundo.

#footer { color: transparent; }

That's not such a good idea. "transparent" runs much the same risk
of cascading badly with other stylesheets, as not specifying it at
all (in detail its cascading properties would be different, but it's
nevertheless the same principle).

Would depend on how many style sheets are used, of course. Should not be
a problem with just one.
It's best if you either specify *both* colours explicitly, or
*neither*, *at each specificity*. It isn't always feasible to do
that in every situation - especially when you have a background image
on an outer element which you want to shine through some inner
element(s) - but it's a good principle to keep to, as far as
possible. This is a *warning*, after all, rather than an actual
error.

Good advice.

--
-bts
-Warning: I brake for lawn deer

[Alan J. Flavell]

On Wed, 15 Feb 2006, Beauregard T. Shagnasty wrote:

Alan J. Flavell wrote:

That's not such a good idea. "transparent" runs much the same risk
of cascading badly with other stylesheets, as not specifying it at
all (in detail its cascading properties would be different, but it's
nevertheless the same principle).

Would depend on how many style sheets are used, of course.

But in the final analysis, the author has no control over whether the
user is applying a user stylesheet or not. That is, after all, one of
the features of *Cascading* style sheets.

Some users might be applying a user stylesheet for accessibility
reasons (e.g due to eyesight problems they prefer to read light
coloured text on a dark ground), so this is a WAI issue too - see the
last sentence of section 9.1
http://www.w3.org/TR/WCAG10-CSS-TECH...color-contrast , although
the sentence omits to explain the background to the advice.

If both participants follow the rule when formulating their respective
stylesheets ("specify either both colours or neither, at each
specificity") then they get the best collective chance of a readable
result, in the face not only of regular CSS cascading, but even in the
face of browser inheritance bugs. The end result of the cascade might
look a bit odd, but at least each text will be presented against a
chosen background colour (assuming that any browser inheritance bugs
would almost certainly affect foreground and background colours
equally).

And this can be done without compromising the appearance for users who
accept the author's stylesheet in its entirety, which is how things
are supposed to work (IMHO, anyway).

If either of them fails to follow the rule, the result *could* be some
black text on a black background (etc.), with the reader being unaware
that they are missing anything. Even if it only ends up with poor
contrast, it's irksome.

[As I said, this agenda isn't always feasible (primarily when
background images are involved), but I'd recommend following the
principle when possible. And not just because it causes the CSS
checker to stop bleating...]

cheers