2009/8/17 Naveen Kumar nav007@gmail.com
2009/8/17 प्रविण सातपुते pravin.d.s@gmail.com
2009/8/16 Naveen Kumar nav007@gmail.com
Currently it's just a keymap which makes use ofthe
facilities provided to us by the iiimf (Internet/Intranet Input Method Framework). But I intend to build more over it, maybe a full fledged mathematical package . And while compiling it I came across a lot of demerits of current Unicode encoding scheme for mathematical text
input.
might be there only one, missing some maths symbol
In the next version I will try to put all possible symbols. Maybe even some additional things.
If I am not wrong TeX/LateX use font based rendering scheme to generate
final
images of the mathematical formulas which are not character symbols.
Unicode has incorporated many symbols already in Unicode, so as far as Character symbol has corresponding Unicode code point and font has corresponding shape included in it, i think it should not create any problem
At some places complicated things cannot be written using just Unicode symbols.
For example when we require any script to act as super script as in symbols for upper limit and lower limit in summation,
or let us say 2^(2^(2^2)) were the 2 must apply 3 levels of superscript styling.
i guess this thing we can handle in font itself, by writing GPOS (glyph positioning rule) rules
Also the symbols like integration and summation must have some kind of place holders to depict these upper and lower limits.
this thing also can done same way, but writing will be something like "lower limit + integration sign + upper limit"
i dont think so unicode willl do anything in this, they can only give integration sign code, with font and some tool we can make it look like real maths handwritten equation.
extendable natural horizontal line for separating numerator and denominator part of a fractional equation. Right now we might have to do away with (/) symbol.
hmm, this is also we can handle in font
i.e when one press 5 then divide sign, change "/ " to horizontal line and next one number will come behind that with GPOS rule
really OT fonts have provide numerous feature to do all these things
The same case with nth root; the horizontal line must cover the equation from the above completely. Right now we do away with (√ ∛√)
again GPOS, but i am really worried about writing sequence it should be standardize for these thing
There are numerous such examples. Some of these demerits can be easily removed while others may require careful thinking. These need to be removed gradually by Unicode consortium, provided some agency works on it. Also the font based mechanisms of latex could be incorporated, but in order for rendering engine to take such decisions, such instructions need to be provided to them. And incorporating specialized code points like \Ux200C and \Ux200D to do these might suffice to some extent. I will try and highlight these things in the white paper. Just give me some time to do just that.
Yep Take your time I think this will be good topic for in any typography conference
Thanks & Regards, ---------------------- Pravin Satpute