Re: Regarding the appearance property

> [Original Message]
> From: Tantek Çelik <tantek@cs.stanford.edu>
>
> > On Sun, 21 Mar 2004 14:05:57 +0100, David Håsäther <hasather@msn.com>
> > wrote:
> > 
> >> Suppose I do something like
> > 
> > snip
> > 
> >> Does this mean that h1 elements will be back to normal again (e.g.
> >> displayed in a bigger font-size, bold and with margins)?
> > 
> > No.  Its appearance is already normal.
>
> Correct.  'appearance:normal' really just means that the appearance
> property has no effect on the element.
>
>
> On 3/21/04 9:20 AM, "Ernest Cline" <ernestcline@mindspring.com> wrote:
>
> > In any case, this property needs rather extensive revision
> > in my opinion before it becomes usable as a standard.
>
> Why?  There are implementers that would disagree.

The definition is woefully imprecise.  Exactly which properties 'appearance'
overrides needs to be defined.  The meaning of "normal" needs to be
defined, as there are several possible interpretations, of which the one
you just gave was one I hadn't even initially considered. (Altho, I do like
it better than my originally preferred interpretation.)  As defined,
it is implementable, but it is definitely not usable as a _standard_.
It is too imprecise, as the draft itself acknowledges.

> > For instance, why should this property be limited to interactive media?
>
> That doesn't make sense.
>
> If you think there are additional media groups that it should apply to,
> please provide a list of them with justifications.

All of the visual media groups.  If someone wants to print out a copy
of an interactive form, then by default, using the appearance property
if it were restricted to just interactive groups would mean that it could
be impossible to duplicate the look of the form when printed.

It may be possible that only implementers of UA's that handle
interactive media may wish to implement the 'appearance' property,
but considered by itself, there in nothing in this property that justifies
restricting this property to just interactive media.  Indeed based on
the description of the property itself as currently given, a restriction
to just the visual media types would be easier to justify.

Received on Monday, 22 March 2004 16:56:02 UTC