[whatwg] Web Forms 2.0 Feedback

[James Graham]

Matthew Thomas wrote:

> On 7 Jan, 2005, at 3:57 AM, James Graham wrote:
>
>>
>> Matthew Thomas wrote:
>>
>>>
>>> Again, I know that <sup> and <sub> are (almost always) used to mean  
>>> something, just like <b> and <i> are. But again, just as with <b> 
>>> and  <i>, *a computer can't tell what you mean*.
>>
>> ...
>> Of course the same is true with, say <li>. A computer can't tell 
>> whether you mean a list of shopping, or a list or a list of links, or 
>> a list of people who have offeneded you in the past month, or ...
>
>
> A computer can tell that it's an item in a list, which is useful by 
> itself <http://labs.google.com/sets>.

Is there anything indicated that uses <li>? It could just use 
coincidence on the same page (indeed, that seems more likely). In any 
case that (in it's present form) is very much on the verge of what I 
would describe as "useful".

> In contrast, a computer can't tell anything at all about <sup> and <sub>.

It can tell that the word containing them isn't to be interpreted as a 
single word which is essential to a useful interpretation of the document.

>
>> ...
>> HTML by it's nature has weak semantics. That means that elements 
>> should conatin some information ("this is a list not a set of 
>> paragraphs", "these characters are superscripted and so not part of a 
>> word") that the UA can use, as far as it is able, to provide an 
>> appropriate interface to the document. It does not mean that every 
>> element has to have a precidely defined meaning in the sense that you 
>> criticise <sup> and <sub> for lacking.
>> ...
>
>
> Where did you get the idea that I was criticizing <sup> and <sub>? 
> They're useful presentational elements, just like <b> and <i>.

You certianly say they're lacking meaning.

But anyway, we've been around this before and I'm not at all sure it's 
getting anywhere.