On Wed, 22 Jul 2009, Michail Vidiassov wrote:

> Dear Hans and All,
>
> On Tue, 21 Jul 2009, Hans Hagen wrote:
>
>>>>> Superscript works well, but not the subscript:
>>>>> 
>>>>> \starttext
>>>>> x³ x₃
>>>>> \stoptext
>>>> Unicode sub/superscripts aren't real sub/superscripts, and there use is
>>>> discouraged. Most fonts will align those subscript glyphs to the base
>>>> line not bellow it, and apparently LM fonts don't have them.
>>> 
>>> In MKIV one could theoretically fix the (virtual) font with subscript
>>> three containing "\sub{3}" *** or "lowered superscript 3" :)
>>> 
>> it's also font dependent and we would en dup with fallback mess etc etc
>> 
>> if the font has the glyph, you can use the unicode char, bu tit is unlikely 
>> that there will be a mapping mechanism
>
> What about an analog of "ucs" LaTeX package,
> when Unicode math subscripts (and other symbols) in TeX file are replaced
> by "\sub{3}" etc., regardless of the font used for typesetting?

Most of the support is already there (except superscripts and subscripts). 
MKII implemented this support using active characters (similar to what ucs 
does). MKIV does this using virtual fonts.

> It will give WYSIWYG-like appearance of the input file on the screen
> in the editor using Unicode font and will not bring the mess deep into 
> ConTeXt?
>
> Is such thing doable?
> (Not making a set of mappings for presonal use, but proper package
> for public consumption?)

For most glyphs the support is already there. Making sure that multiple glyphs
are combined correctly is tricier. We can have an input like $ŵ¹$ and it is not
always easy to determine if it should be translated to ${\hat w}^1$ or $\hat
{w^1}$.

Microsoft implmented something like this in newer versions of Word. See the
developer's blog for more detials.
http://blogs.msdn.com/murrays/archive/2009/05/07/entering-math-via-the-linear-format.aspx

Aditya