RE: Fwd: Re: (SC2WG2.609) New contribution N2705

From: Peter Constable (
Date: Wed Feb 18 2004 - 10:47:03 EST

  • Next message: Antoine Leca: "Re: Devanagari Letter Short A"

    > From: []
    On Behalf

    > Why couldn't there exist a combining subscript modifier, added after a
    > character, to make it look like a subscript, and that could be given a
    > combining class 255?

    Because this is Unicode and not CourtyardCode?

    This is *not* a good idea.

    > So that's
    > something which can be handled in fonts.

    I need to create *ligatures* for thousands of possible base-subscript
    pairs in all of my fonts? I've got to decide how I should display < a,
    combining acute, combining subscript modifier >, possibly adding
    thousands more base-subscript-combining mark ligatures into my font?
    I've got to decide how I should display an Arabic sequence like < seen,
    seen, combining subscript modifier, seen > (what does or doesn't
    connect?)? And Devanagari < ka, virama, combining subscript modifier,
    ssa >? And < c, ZWJ, t, combining subscript modifier > or < c, combining
    subscript modifier, ZWJ, t >? I've got to decide how I should display <
    U+00B2 superscript two, combining subscript modifier > and < U+2081
    subscript one, combining subscript modifier >?

    Not to mention that I've got to tell our app developers they have to
    figure out whether users expect < subscript one > and < digit one,
    combining subscript modifier > to sort together and be matched in
    searches? And whether < digit one, digit two, combining subscript
    modifier > is interpreted as having a numeric value of 12 or 1?

    I don't think this rocket will make it off the pad, let alone past the

    Peter Constable
    Globalization Infrastructure and Font Technologies
    Microsoft Windows Division

    This archive was generated by hypermail 2.1.5 : Wed Feb 18 2004 - 11:41:19 EST