[whatwg] Improve select required
whatwg-2010-04 at jessemccarthy.net
Mon Sep 13 11:20:43 PDT 2010
On Monday 13 Sep 2010 Mounir Lamouri wrote:
> 2. Introduce a placeholder boolean attribute for <option> and select
elements will suffer from being missing if there are no options selected
or selected options all have the placeholder attribute set.
I also wonder if that would be more straighforward. I suggested / inquired
about that when this was being discussed before, but the discussion petered
out and no one said anything about it:
(See last paragraph.)
> last solution isn't [backward compatible] (ie. the placeholder text
In my opinion, something that relies on scripting isn't a viable solution.
> Solution 3, seems to be the nicest and is consistent with the other part
of the specifications (placeholder would be used in <input> and <select>).
That's not viable now because it's not backward compatible, but if people
think that would be the ideal solution in the future, would it feasible to
specify and implement the following authoring requirements and user agent
* @placeholder for SELECT (as in your #3)
* @placeholder as a boolean attribute on OPTION (just as a crutch for
* @placeholder on OPTION is only valid when @placeholder is also used on the
SELECT (and perhaps only if the OPTION's textContent matches the value of
the SELECT's @placeholder attribute?)
* Although @placeholder on OPTION is valid / tolerated under those
circumstances, conforming UAs must ignore those OPTIONs entirely and use the
value of @placeholder on the SELECT
If that's doable, then implementors could implement the behavior for
@placeholder on SELECT now and authors could author documents that use the
new mechanism and preserve the status-quo for HTML 4 user agents -- an
initial "label" OPTION. When authors stop caring about HTML 4 UAs, they can
just stop including an OPTION with @placeholder.
More information about the whatwg