Re: Styling Controls (was: native elements versus scripted)

On Wed, Apr 7, 2010 at 12:17 AM, Doug Schepers <schepers@w3.org> wrote:
> Hi, Shelley-
>
> Shelley Powers wrote (on 4/6/10 11:21 PM):
>>
>> On Tue, Apr 6, 2010 at 9:58 PM, Jonas Sicking<jonas@sicking.cc>  wrote:
>>>
>>>  On Tue, Apr 6, 2010 at 7:33 PM, Shelley Powers<shelley.just@gmail.com>
>>>  wrote:
>>>>
>>>>  If there was interest in doing this before, why is only now people are
>>>>  asking these questions?
>>>
>>>  If I understand your question right, in large part nothing has
>>>  happened in this area in the past because W3C stopped development of
>>>  HTML after the release of HTML4.
>>>
>>
>> Jonas, I know the history of the WhatWG effort. It started with Web
>> Forms 2.0. These web elements were not added to the spec a couple of
>> months ago. They've been in the spec for years. The datetime input
>> element type was added to Web Forms 2.0 by a group of Opera, Mozilla,
>> and Apple folks in a private email list over six years ago.
>>
>> So, folks knew that these issues were going to arise at some point.
>> And since these elements didn't arise in HTML4, I'm not sure why it
>> would be pertinent to the current situation.
>
> I suspect the reason it hasn't been discussed before is that there are an
> unlimited number of things we can do to improve the Web platform --a list of
> things that grows and changes rapidly-- and a limited number of people to
> specify and implement them... and those people often disagree about what the
> right thing to do is.  Sometimes you have to step back and take a holistic
> look at the platform and identify what's missing, and what has changed in
> priority.  Though I differ with your conclusions, I think your questioning
> of these new form controls has helped highlight an important issue... how do
> we style these elements and their pseudo-elemental components?
>
> I think that a careful examination of the common constituents of these
> controls would give a lot of power to designers while still allowing for
> innovation by browser vendors in rendering them.  For example, you showed 2
> different styles of slider, one with tickmarks and a rectangular thumb and
> the other a simple bar with a round thumb; a designer could still change the
> color of the thumb and bar (and maybe things like rounding of corners) and
> specify the color of tickmarks if they happen to be present on that platform
> (or to hide them if they so choose); I alluded to this in my earlier email
> that kicked off this topic of styling native controls [1]... if no tickmarks
> are normally rendered, that pseudo-element rule simply won't be applied, and
> that's okay.
>
> It may have occurred to me more readily because, like a lot of people in the
> SVG developer community, I've been rolling my own SVG controls for a decade
> or so, in several different libraries, where each component (slider bar,
> thumb, tickmark, labels, etc.) is a distinct element, and thus easily
> styled.  People who create HTML equivalents out of <div>, <p>, etc. have a
> similar experience, I imagine.  I see ARIA as a real boon for these custom
> controls, especially in SVG where we never intend to have native form
> controls, but I see equal need for CSS to afford native controls the same
> kind of stylability (and I'm glad to see that some browser implementers and
> CSS WG folks are amenable to that).
>
> [1] http://lists.w3.org/Archives/Public/www-style/2010Apr/0076.html
>

I agree on the benefit of ARIA. I've become rather a fan. I don't
think of it as accessibility, I think of it as "rendering semantics".

I'm not sure if the same concerns that have been expressed in the
past, and codified in the conformance section of the CSS 2.1
specification are different now. For instance, I don't particularly
think we should be able to style buttons or selection elements beyond
what we already can. They are simple, uncomplicated controls and don't
need to be gussied up. Doing so does have a potential impact on
usability.

I am concerned, though, that all of a sudden, there's this intense
interest to style form elements, all occurring at the same time that
we we're really challenging both the relevance and usefulness of new,
complex, and generally unimplemented form elements.

However, I'm not a member of the CSS WG, so have no say, one way or
another. Perhaps you or another member of the CSS WG can link ongoing
discussions on this new interest in form element styling in CSS back
to this group, so we can keep informed.

Thanks

> Regards-
> -Doug Schepers
> W3C Team Contact, SVG and WebApps WGs
>

Shelley

Received on Wednesday, 7 April 2010 05:41:44 UTC