Re: The title Attribute (Was: conflation of issues or convergence of interests?)

At 09:19 +0100 UTC, on 2007-07-31, Smylers wrote:

> Sander Tekelenburg writes:
>
>> Empty title attributes are author mistakes (and should be warned for
>> by HTML checkers).
>
> Why?

Well, I was speaking in the context of authors using it as a trick to
suppress @alt tooltips. Simply the fact that HTML describes documents, not
behaviour, means that that is an author mistake. Another thought behind my
dismissal was that an empty title attribute is meaningless, yet might very
well trigger a UA to present the empty value (after all, nothing in HTML 4
says UAs must treat empty @titles specially), which would probably be harmful
(at least in the sense of confusing users).

However:

> For comparison, the spec currently gives empty title attributes a
> meaning, overriding a parent element's title attribute (with nothing):
>
>   If [title] is omitted from an element, then it implies that the title
>   attribute of the nearest ancestor with a title attribute set is also
>   relevant to this element.  ... Setting the attribute to the empty
>   string indicates that the element has no advisory information.

Gosh...

I hadn't yet noticed that. Is this needed? useful? harmful?

> Are you claiming that it is always a mistake for an author to decide
> that a parent element's title doesn't apply but that the current element
> doesn't need a title?

You surprised me with this finding. I'd need to think about the answer. I
assume there is a rationale for this change from HTML 4. If someone could
post an example of a case where an empty title attribute would be
needed/useful, that would certainly be helpful.

> Or do you think it's wrong for title attributes to be inherited like
> this, and you'd rather that omitting the title attribute always
> indicated that no title is relevant?

I see the title attribute as something that applies to the element and its
children. Simply because it applies to the element. Obviously there are
situatoins where a child element would need its own title attribute, which I
would consider a sort of "detailed refining". Not an 'override' -- the
parent's title attribute still applies as well. I can't realy think of why an
author would need to 'disable' a parent element's title attribute. Seems to
me that when that is necessary, that child doesn't belong to that parent at
all in the first place.

<parent car=volvo>
	<child>
	<child car=trabant>
	<child>
</parent>

Why would any child need to be marked 'not volvo'?

But again, someone must have thought of this, so they can probably give
examples of situations where the HTML5 use and meaning of an empty title
attribute would fulfill a need (and be the best solution, and not be harmful
;)).


-- 
Sander Tekelenburg
The Web Repair Initiative: <http://webrepair.org/>

Received on Tuesday, 31 July 2007 10:29:36 UTC