On Thu, 4 May 2017 23:13:08 +0000 Michael Bear via Unicode <unicode@unicode.org> wrote:
> I plan to do everything in the plane EXCEPT for the surrogates, which > you’re not supposed to encode in fonts anyway, which leaves room for > about 2,048 more glyphs for OpenType features. There are, if I avoided double counting errors, 56,251 assigned characters in the BMP in Unicode 10.0.0. There are 1008 standardised variation sequences, all in the BMP. Indic scripts require more glyphs than they have characters - usually at least twice as many. You have the read chapter on Devanagari, haven't you? Richard.