[whatwg] Web Forms 2.0 Feedback

Ian Hickson ian at hixie.ch
Wed Apr 13 04:01:25 PDT 2005

(I hope y'all don't mind me replying to all your e-mails out of order. I'm 
basically going down the spec one element at a time and when I come across 
one that someone has discussed in the past, I reply to those e-mails.)

On Fri, 7 Jan 2005, Matthew Thomas wrote:
> On 7 Jan, 2005, at 5:58 AM, Ian Hickson wrote:
> > ...
> > For <sub>, the ideal aural rendering depends on the context, but humans
> > are adept at interpreting things based on context and you could probably get
> > away with rendering sub by simply prefixing its contents with the syllable
> > "sub", as in "H sub-two O" for "H<sub>2</sub>O".
> > ...
> > the fact that you can use the element to sensibly change the aural rendering
> > suggests to be that it is semantic enough to be kept in HTML.
> Except that "sub" is merely (an abbreviation of) a description of the
> typographical presentation! You might just as well say that H<font
> color="green">2</font>O is semantic because it could be pronounced as "H
> green-two O".

But it wouldn't be, and that's rather my point. I rarely hear people 
calling out the fact that something is bold or italics or red when they 
are reading text out. I _do_ hear them say "sub-" this and "super-" that.

> > It's not ideal, and for a better aural rendering you would use a 
> > speech-capable UA that supported ChemML, MathML, or another more 
> > appropriate standard language natively, and pass content to it using 
> > the appropriate domain-specific language.
> FWIW, in the case of ChemML that wouldn't help -- as far as I know 
> (having just consulted the household science teacher), there is no 
> standard way of pronouncing chemical formulas so as to distinguish 
> subscripts even from normal numbers. So using aural rendering to decide 
> whether an element is semantic wouldn't work in that case.

There's no standard rendering, but there are conventions that a UA author 
can use.

There are several use cases I can see for <sub> and <sup> that I consider 
to be semantic and not presentational. However, it seems very much to be 
on the fine line between the two, and I am interested in hearing of ideas 
that would more clearly move <sub>/<sup> or new elements into the semantic 

Use cases:

   In french, parts of words (abbreviations, usually) are always 
   superscripted. For example, the "lle" in "Mlle".

   In chemistry, number of atoms, atomic weights, charge, etc, are
   superscripted and subscripted.

   In variables names parts are often subscripted to indicate a specific 
   variable in a family of variables.

   In maths, superscripts are used for powers.

And of course in lots of other fields there are specific uses for them 
too. I propose to try to address as many of possible of these use cases in 
specific ways. For example, "Mlle" would be:


...and variable subscripts would be:


I'm not sure how to deal with the chemistry case. We don't really have an 
element for anything like chemical formulas. For maths, MathML comes to 
mind, although only for the XML serialisation.

Any other cases? Any suggestions?

Ian Hickson               U+1047E                )\._.,--....,'``.    fL
http://ln.hixie.ch/       U+263A                /,   _.. \   _\  ;`._ ,.
Things that are impossible just take longer.   `._.-(,_..'--(,_..'`-.;.'

More information about the whatwg mailing list