[whatwg] Suggestion: Implementation of Tabbed Forms

Lachlan Hunt wrote:
>   Fair enough.  Perhaps a name that better represents what Ian Hickson 
> calls them: mutually exclusive sections, which, BTW, is quite a good 
> descriptive name.

    I have no problem using the term "section", although I don't know it 
that will work as an element name, as I will explain a little later in 
this message.

>>  The sections may even be indistinguishable from each other except for 
>> the fact that they are in different sections...
> 
>   This has confused me, I'm don't understand what you mean.

    I've actually worked on applications where you have tabs like 
"Chromatography 1", "Chromatography 2", et cetera, where the contents 
where identical. Different stages of a process may still need the same 
set of configuration variables.

>> ...still a problem. <category> and <categoryset> elements
>> would degrade to nothing in IE 6.0... styling would not show up.
> 
>   So what?  That will happen with any new element introduced. 
> Internet-explorer-operability is only needed to the extent that IE users 
> can still access the content and use the application.  Also, it seems 
> likely that scripting may be used to simulate behaviors and styling in 
> IE and other non-WF2 user agents...  For example, I'm working on a 
> script right now that can generate a date picker control, for use with 
> <input type="datetime"/>, as well as parsing and generating ISO 8601 
> compliant dates (since ECMAScript doesn't natively support them) ? I 
> might have that done in a few weeks, if I can find some spare time to 
> fit everything in.

    The difference is that, in IE without such scripting support, the 
date control degrades into a text input, so there is at very least a 
control there to input the date into. With elements that IE doesn't 
recognize, IE with scripting disables would render NOTHING, regardless 
of what you put into CSS. The styling for unknown elements is ignored 
entirely. (Mozilla allows such styling, on the other hand. I haven't 
tried it on Opera or Safari.)

>> <div type="groupset" id="firefox-options">
>>   <div type="group">
>>     ...
> 
>   No, as I [1], and Anne van Kesteren [2] wrote previously, the type 
> attribute should not be used for anything except specifying the content 
> type of an external resource and the control type for input elements.

    It it the |type| attribute you object to, or using an attribute for 
declaring semantic relationship between the children of the <div> 
element? I was attempting to create a solution that could theoretically 
degrade into your earlier tabs example.

    Note that there is now way to know if a WF2 emulation layer will be 
available for every non-WF2 UA. It  may actually turn out that only an 
IE-specific emulation layer will be made, since the next three biggest 
browser developers are going to include native support. Heck, Safari is 
based on Konqueror, so you have the two biggest Linux browsers 
supporting it natively right there. That means people forced to live 
with older browsers will have to settle for degraded markup.

Received on Wednesday, 30 June 2004 08:49:39 UTC