Re: CSS Validator Question

i agree.

so what do you think the validator should do?  I
personally think it should stick to syntactical stuff.
Comparing it to a compiler, this color scheme error
would be more along the lines of a run-time error,
whereas the validator should be be picking up
compile-time errors.

i think sooner or later you have to just admit that
a validator is nice and helpful, but just because your
page passed the validator's tests, it does not mean
your page is OK.  Maybe your code is OK, but that
code is only as good as the idea behind it, and let's
face it, the web is full of bad ideas.



~b



* this page validated by Human Brain v.1.0 beta






----- Original Message -----
From: "Ernest Cline" <ernestcline@mindspring.com>
To: "W3C CSS List" <www-style@w3.org>
Sent: Friday, September 12, 2003 10:22 PM
Subject: Re: CSS Validator Question


>
> > From: b <mighty_joe_young@yifan.net>
> >
> > You shouldn't have to declare a background color EVERYTIME you declare a
> > foreground color, so long as some cascading is occuring. ("huh"?)
> >
> > Take this example:
> >
> > DIV {
> >   padding:1em;
> >   border:solid 1px #999;
> >   background-color:#339;
> >   color:#999;
> > }
> >
> > DIV A {
> >   color:#FFF;
> > }
> >
> >
> > This produces silver text on a dark blue background, and white text on
> > anchors.  Tacky? Sure.  Visible? Yes.
>
> Except that if there is a
>
> <div>
>   <p>
>     <a>
>
> And another stylesheet has the rule:
>
> P {
>   background:color:white;
>   color:black;
> }
>
> and no rule for DIV A or P A, then the result is white text on a white
> background.
> (Doesn't matter which is the user sheet and which is the author.)
>
> That is one reason why if a validator is going to warn about this it,
> then it shouldn't make any inheritance assumptions.
>
>
>

Received on Friday, 12 September 2003 23:54:18 UTC