[whatwg] Re: several messages about Web Forms 2

Ian Hickson ian at hixie.ch
Thu Jul 21 17:29:56 PDT 2005


On Wed, 20 Jul 2005, Anne van Kesteren wrote:
>
> Any chance Web Forms 2 is going to say something about:
> 
> <!DOCTYPE html>
> <title>Default value for size DOM attribute from the SELECT element</title>
> <p><select><option>Test</select>
> <script>
>  alert(document.getElementsByTagName('select')[0].size);
> </script>
> 
> Firefox 1.8b3: -1.

But if you actually _set_ the size="" attribute to -1 it turns into 0, go 
figure.

I hadn't said anything about size="" in WF2 because I can't work out what 
to say. How do <select size="3"> and <select> differ for, e.g., a speech 
browser? Yet clearly there is an important difference; The difference 
between a drop-down select box and a list box is important from more than 
just a presentational point of view.

I've put in some text about size's default value for now, but this will 
clearly have to be revisited in WA1.


On Thu, 21 Jul 2005 fora at annevankesteren.nl wrote:
> > >
> > > <xh:select xmlns:xh="http://www.w3.org/1999/xhtml">
> > >  <xh:option>FAIL</xh:option>
> > >  <foo xmlns="tag:example.org,2005:/test">
> > >   <xh:option selected="selected">PASS</xh:option>
> > >  </foo>
> > > </xh:select>
> >
> > It should do exactly the same as if that markup was included directly 
> > inside an XHTML file; the data="" attribute is not relevant here.
> 
> Same for a text/html file?

Well, HTML has no namespaces, but had you meant with DOM calls creating 
the above structure, yes.


> Sorry for not being more specific the first time. I'm especially 
> wondering how the namespace handling should go inside text/html 
> documents. I know that text/html documents should basically be 
> considered to be inside the XHTML namespace but should do be treated as 
> XML documents as well? From a DOM point of view?

From a DOM point of view, there is no difference between XML and HTML. 
They're all just nodes.


On Thu, 21 Jul 2005, Anne van Kesteren wrote:
>
> What exactly should happen here:
>  <http://annevankesteren.nl/test/html/wf2/select/007>

FAIL. Or at least I assume so. I haven't really changed anything from the 
DOM2 HTML specs as far as <select> goes (including trying to clarify 
things).

I've added some text defining it.


On Thu, 21 Jul 2005, Anne van Kesteren wrote:
>
> [whitespace trimming on the type attribute]
> It should basically not happen at all. It appears that no browser has
> implemented the 'may' from HTML4 and therefore we can now say browsers must not
> trim.

Done.


On Thu, 21 Jul 2005, Christoph Päper wrote:
> >
> > and therefore we can now say browsers must not trim.
> 
> IMO trimming is still desired behavior (and fairly easy to do), 
> therefore "must not" is way too strong.

We want interoperability (as in, all browsers interpreting the same markup 
to mean the same thing). Therefore it has to be either "must" (in which 
case, according to Anne's findings, everyone is non-conformant), or "must 
not". What's the advantage of "must not"?


On Thu, 21 Jul 2005, Kornel Lesinski wrote:
> 
> That conflicts with XML rules:
> 
> "If the attribute type is not CDATA, then the XML processor must further 
> process the normalized attribute value by discarding any leading and 
> trailing space (#x20) characters, and by replacing sequences of space 
> (#x20) characters by a single space (#x20) character."

The attributes are all CDATA (since there is no DTD).

But I've left a note to the effect that XML could override this.


On Thu, 21 Jul 2005, Anne van Kesteren wrote:
> From:
>  <http://whatwg.org/specs/web-forms/current-work/#methodAndEnctypes>
> # The value of the enctype attribute must be dispatched using a
> # case-insensitive literal comparison.
>
> What exactly does this mean, in terms of Unicode?

Defined (thanks to Bjoern for some pointers).

-- 
Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'


More information about the whatwg mailing list