[uf-discuss] a very early draft proposal hTagcloud

[Ben Ward]

On 20 Sep 2006, at 02:23, Chris Messina wrote:
> I think what you need to define are ways to express relativety -- and
> that <strong>, <em>, <big> and <small> can help in indicating those
> relationships with styles turned off. So for example, the very
> smallest size might always have <em> surrounding the <small> tags...
> so that there's a lower limit. You could use <em> or <strong> around
> the <big> tags to express the upper limit.

I'm not sure about using BIG and SMALL. We start getting into the  
territory of redefining elements (adding semantic values to those  
presentational elements[1]) which is something the WHATWG are already  
doing with HTML5. Mainly, they're redefining SMALL for ‘small print’  
in documents [2], which conflicts directly with using SMALL in tag  
clouds.

Personally I quite like the nested behaviour. I accept that is also  
unspecified implied semantics, but it doesn't seem unreasonable… that  
said, you can get four levels of tag cloud without nesting of the  
same element (by using STRONG as well), e.g:

<ol class="hTagcloud">
   <li>A Level One</li>
   <li>B Level One</li>
   <li><em>C Level Two</em></li>
   <li><strong>D Level Three</strong></li>
   <li><strong><em>E Level Four</em></strong></li>
   <li><strong>F Level Three</strong></li>
   <li>Level One</li>
</ol>

Would four levels be enough?

Ben

[1] http://www.w3.org/TR/REC-html40/present/graphics.html#h-15.2.1
[2] http://whatwg.org/specs/web-apps/current-work/#the-small