Aw: Re: Missing Latin superscript lowercase letters

[Giacomo Catenazzi]

On 23 Mar 2023 15:55, Marius Spix via Unicode wrote:
> In TeX and MathML, U+005E CIRCUMFLEX ACCENT ^ is used for superscript 
> and U+005F LOW LINE _ for subscript. This also allows power towers like 
> 2^(2^2), which are not possible with the existing Unicode characters. 
> This notation is recognized by mathmaticians, physicists and chemists 
> and widely accepted.

> In some programming languages, e. g. Java or C++, ^ is used for the XOR 
> operation and _ for digit grouping, but that does not matter here, 
> because the context is always the decisive factor.

> Many modern fonts have support for auto-alignment of digits in 
> combination with U+2044 FRACTION SLASH ⁄ like in that example: 13⁄37. So 
> it may be possible to design a font with special handling for ^ and _.

Let's be honest: most people which are asking for superscript letters 
are not interested in display in mathematics.

In any case /FRACTION SLASH/ can display some fractions, not all fonts 
have good support for it (e.g. on your mail I see it nice, but in the 
quoted text above I see it very ugly. But such formatting not only 
depends on font, but also on user preferences (and settings). Like 
tabular numbers (also often included in fonts), it may not be enabled by 
default, or it should be explicitly enabled e.g. on tables). Unicode do 
not support such presentation settings. So a FRACTION SLASH is just a 
solution for one specific case. Note: there is no delimiter, so usually 
fonts support it only for numbers, and maybe with limited number of 
digits. Maths is more then nubers. An other reasons to have maths in 
markup language.

Also it is very annoying to type Unicode symbols not on the restricted 
number of keys in normal keyboards. We learn it also from computer 
languages. On earliest days computers languages had many symbols because 
every operations "needed" own symbols. Guess what? Now every modern 
computer languages uses practically only ASCII characters. 
Practicability is better then a ideal system few people uses.

In any case, to display true maths, we need a specialised engine (and 
fonts). We are far from having current shaping engines (and fonts) to 
display maths in a nice way. (and personally I prefer that developers of 
shaping engines will works on improving the actual engine and fonts for 
human languages, before to go on such specialised field (which we have 
already good tools).

Superscript letters can be done with current fonts and current shaping 
engines and many markup languages, so any discussion (and new 
characters) are distractions which do not direct us on a true Unicode 
mathematical typesetting (not a goal, like musical notation). And it 
will make things worst: searching engines must have to interpret 
everything. Speech synthesis will become much more complex (and it 
should understand where it is maths, chemistry or units: you will need 
to spell them differently. And probably many other unforeseen problems.

ciao
	cate