[pavel_lishin]

I mean text as in the platonic ideal of "c" and "с". Just because they look the same, does not make them the same character. If we're going to be encoding characters that happen to have pixel-identical renderings in certain fonts, the next logical step is to encode identical letters that look different in different fonts or writing styles as separate code points as well - for example, the English letter "g" is a fucking orthographic nightmare.

[kps]

Imagine if, say, English people normally wrote an open ‘g’ and French normally wrote a looped ‘g’, and you have the essence of the Han Unification debates.