Re: WD-CSS21-20020802 section 8, "Box model", substantive comments

Ian Hickson wrote to <www-style@w3.org> on 27 February 2003 in
'Re: WD-CSS21-20020802 section 8, "Box model", substantive comments'
(<mid:Pine.LNX.4.50.0302270536500.14072-100000@dhalsim.dreamhost.com>):

> If you can come up with criteria that are suitable, it is likely that they
> will be used. However, personally I consider the current text suitable,
> because in practice humiliation is just as good an incentive to compliance
> as is watertight normative text...

Humiliation may be a good incentive (I'm not sure that it is), but not, in
this case, to compliance. It is the normative text which defines compliance.
Neither prevailing wisdom, nor consensus among list members, nor collective
consciousness has any bearing in the definition of compliance once
published. So the humiliation would be an incentive to something above and
beyond compliance.

I think that what is bothering me is not so much the prospect of
misimplementations; I've seen plenty during my short love affair with CSS,
and some more won't really hurt me. No, what is bothering me is the prospect
that a vendor could release a decrepit implementation, call it compliant,
and be correct.

On the bright side, I believe that the fix is easy to specify. Please
consider the following passage, written as HTML, for inclusion.

<p>User agents may diverge from the border model described when rendering
<abbr title="Hypertext Markup Language">HTML</abbr> and <abbr
title="Extensible Hypertext Markup Language">XHTML</abbr> elements of the
following types.</p>
<ul>
 <li>a</li>
 <li>applet</li>
 <li>area</li>
 <li>button</li>
 <li>fieldset</li>
 <li>frame</li>
 <li>frameset</li>
 <li>hr</li>
 <li>iframe</li>
 <li>img</li>
 <li>input</li>
 <li>label</li>
 <li>legend</li>
 <li>link</li>
 <li>map</li>
 <li>object</li>
 <li>optgroup</li>
 <li>option</li>
 <li>select</li>
 <li>textarea</li>
</ul>
<p>In these cases, there may be multiple borders on a given side, borders
may be curved, the user agent may omit borders altogether, and so forth. Any
border treatment is considered conformant for these element types.</p>

Received on Friday, 28 February 2003 03:49:59 UTC