W3C home > Mailing lists > Public > www-style@w3.org > April 2003

Re: Scrollbars In The Box Model

From: Ian Hickson <ian@hixie.ch>
Date: Fri, 11 Apr 2003 10:07:39 -0700 (PDT)
To: Dylan Schiemann <dylans@yahoo.com>
Cc: "www-style@w3.org" <www-style@w3.org>
Message-ID: <Pine.LNX.4.50.0304110934240.17269-100000@tavis.dreamhost.com>

On Tue, 8 Apr 2003, Dylan Schiemann wrote:
> I feel that the working group is inconsistent in deciding when something
> should be non-normative, and when something should be added to the
> specification to reflect existing coding practice.

Committee-driven designs often feel inconsistent. It's to be expected
given the development model.

> In the case of X11 named colors, it was argued that the working group
> was including named colors to reflect existing practices.  I and many
> others argued against including X11 named colors, or at least
> deprecating them, because they are a poor, inconsistent collection of
> names.  As far as I can tell, they are still part of the CSS colors
> module.

I am not 100% familiar with the issues, but I believe that the SVG group
requested these colours stay in the spec without deprecation. In any case,
there is no amiguity in the definition of these SVG colours, so there
isn't really a parallel with scrollbars.

> While I do respect Ian's rationale for not specifying where scrollbars
> lie in the box model, I find this to be a more necessary thing to have
> in a specification than named colors.

Scrolling mechanisms vary *wildly*. There are scrollbars, panning
mechanisms, purely input-based systems, buttons placed on the overflow
edges (as seen in long menus on Windows), automatic marquee, etc.

It isn't the place of the CSS working group to specify what scrolling
mechanism should be used.

I do agree that it would be beneficial to explicitly specify that in the
case where a UI widget is inserted on an edge of the element, it should be
inserted between the border and the padding, though.

> Certainly the scrollbar mechanism in browsers is more well-conceived
> than X11 colors with names like Dodger Blue

Whether something is well conceived is not as important as whether it is
implementable when it comes to making it normative.

> I'm just looking for some consistency here.  The working group should be
> looking to write a spec that reflects existing practices, or a spec that
> provides a well-thought out vision for future development, rather than
> both in a nonconsistent manner.

Unfortunately, while that would be ideal, it isn't always possible when
you are trying to do a balancing act between several groups of people each
with totally different requirements.

Ian Hickson                                      )\._.,--....,'``.    fL
"meow"                                          /,   _.. \   _\  ;`._ ,.
http://index.hixie.ch/                         `._.-(,_..'--(,_..'`-.;.'
Received on Friday, 11 April 2003 13:07:43 UTC

This archive was generated by hypermail 2.3.1 : Monday, 2 May 2016 14:27:06 UTC