Re: [selectors4] :empty psuedo class and input fields

Schalk,

I totally understand how it's currently supposed to work(yes my previous
statement about that the spec was not definite was wrong), I simply do not
think it's intuitive that :empty does not check for text in elements which
cannot contain children(inputs)

Regarding that CSS is not event based I beg to disagree, I will not go
through all possible events that are already supported, but :hover and
:focus both respond to mouse events. Why should keyboard events be any
different?



Stefan Wallin
================================
+46 (0) 709-529 036 || stefan.w@festiz.com
http://www.stefan-wallin.se
http://twitter.com/Stefan_Wallin


On Mon, Mar 12, 2012 at 6:37 PM, Schalk Neethling <sneethling@mozilla.com>wrote:

> Stefan,
>
> Remember, whether the input field has text inside it or not, checking for
> childNodes on the input element will always give you an empty array as the
> text is not a child of the element, hence you will find that your first and
> second input elements will have a dashed border and not one solid and one
> dashed, as you might be expecting.
>
> NOTE: This is true if you only specified border color for .toc
> input[value=""] and border style as well. For example change your CSS to
> this:
>
> .toc input[value=""] {
>    border-color: red;
> }
>
> Also, even after adding text to the second input element the border will
> remain red as CSS is not even based and will not be triggered based on a
> change to the input element. It is simply going to style all empty input
> elements a certain way when the document is rendered and that will remain
> that way unless you
> 1) change the style rules at run time
> 2) remove the rule and apply a new one
>
> So, unless there is concrete support in CSS for reacting to changes in DOM
> elements such as change events on input fields, you are going to have to
> combine this with a little DOM manipulation.
>
> Kind Regards,
> Schalk
>
> ----- Original Message -----
> From: "Stefan Wallin" <Stefan.W@festiz.com>
> To: "Schalk Neethling" <sneethling@mozilla.com>
> Cc: "Mounir Lamouri" <mounir@lamouri.fr>, www-style@w3.org
> Sent: Monday, March 12, 2012 5:18:45 PM
> Subject: Re: [selectors4] :empty psuedo class and input fields
>
> Thanks, Schalk.
>
>
> You are true in that assumption. There is a use case where one would want
> to visibly show which elements are not filled in differently from those
> that are required. The naming of the :empty psuedo selector has a very good
> semantic meaning for this and it makes sense for it to be used like this. I
> can envision a lot of uneducated developers to make bad assumptions based
> on how the spec is written. I'm pretty new at wearing specification reading
> glasses, but imho the spec does at this spot say neither against or for
> this functionality.
>
>
> If you need it specified in code I've written a small test here.
> http://lab.festiz.com/psuedo-empty/
>
>
>
>
>
> Stefan Wallin
> ================================
> +46 (0) 709-529 036 || stefan.w@festiz.com
> http://www.stefan-wallin.se
> http://twitter.com/Stefan_Wallin
>
>
>
> On Mon, Mar 12, 2012 at 3:43 PM, Schalk Neethling < sneethling@mozilla.com> wrote:
>
>
> Mounir,
>
> I believe what Stefan wants is to style fields that are empty, but not
> necessarily required, different so input:required:invalid is not going to
> work in this situation.
>
> Regards,
> Schalk
>
>
>
> ----- Original Message -----
> From: "Mounir Lamouri" < mounir@lamouri.fr >
> To: www-style@w3.org
> Sent: Monday, March 12, 2012 4:37:19 PM
> Subject: Re: [selectors4] :empty psuedo class and input fields
>
> On 03/12/2012 11:03 AM, Stefan Wallin wrote:
> > This strikes a chord with me, I'd really like to see the possibility to
> > style input fields that do not fit to be classed as required but are
> > still empty to be styled differently. Today this has to be done with DOM
> > scripting, neither input[value=""] or input:empty works as I think it
> > should work.
>
> This use case can already be solved like this:
> input:required:invalid {
> }
>
> Assuming you are using the required attribute as specified in HTML.
>
> Regarding the general proposal. That might be useful assuming there are
> use cases different from the one above.
>
> --
> Mounir
>
>
>
>