[whatwg] rel/rev for <form> ?

[Matthew Raymond]

[Note: Quotation reordered for clarity.]

> On Nov 8, 2005, at 12:30 AM, Anne van Kesteren wrote:
>>The use of 'class' for presentation is wrong anyway (and hopefully
>>obsoleted in HTML 5). And yes, although it is named incorrectly,
>>the attribute can take multiple, space-separated, values.
>>
>>--
>>Anne van Kesteren
>><http://annevankesteren.nl/>

David Hyatt wrote:
> The class attribute! So efficient it must be wrong! :)

   Yeah, either Anne accidently used the wrong wording or he's lost
touch with reality. The HTML 4.01 spec clearly says classes can be used
as a target for CSS presentation:

| As a style sheet selector (when an author wishes to assign style
| information to a set of elements).

   It also says this just below that:

| For general purpose processing by user agents.

   Note, however, that this is an obvious callback to the "general
purpose processing" that the |id| attribute allows:

| For general purpose processing by user agents (e.g. for identifying
| fields when extracting data from HTML pages into a database,
| translating HTML documents into other formats, etc.).

   This may allow for microformats (although it's more clearly designed
as a way to allow search engines to use |class| as a form of metadata in
the processing of HTML). More importantly, though, is that it says web
authors can go ahead and use |class| to describe the function of a
<form> under HTML 4.01, which search engines can later process. So we
really don't need to redefine |class| for such a role, since it already
supports it.

   (Interestingly enough, if HTML 4.01 allows |class| to be used for
microformats, it must logically also allow the use of |id| for
microformats.)

   I think microformats are cool, but I completely oppose the definition
of any microformat, or dependency on specific microformats, as part of a
WHATWG specification. I consider microformats to be out-of-scope. To me,
they are a way of adding functionality on top of HTML, not a way to add
functionality to HTML. If HTML truly needs to support something, that
support needs too be added as markup, not as a series of predefined
class names.

   (This is why I oppose Ian's current <calendar> and <card> ideas in
the WA1 spec. They're two elements that depend entirely on microformats.
As some might recall, I've already proposed an element-based solution.[1])

   BTW, why hasn't anyone brought up the XHTML2 |role| attribute? It
would seem to be a good fit in this discussion.


[1]
http://listserver.dreamhost.com/pipermail/whatwg-whatwg.org/2005-February/003111.html