Re: U+0345 COMBINING GREEK YPOGEGRAMMENI not usable in other scripts as "hook below"?

From: John Hudson (
Date: Fri Jun 16 2006 - 22:13:58 CDT

  • Next message: Alexej Kryukov: "Re: U+0345 COMBINING GREEK YPOGEGRAMMENI not usable in other scripts as "hook below"?"

    Kent Karlsson wrote:

    > I think that all these changes of subscript to adscript *by the font*
    > is a bad idea. Such changes should be done at the character level,
    > not the glyph level.

    > (And, if I understood right, uppercasing a lowercase (Greek) letter with
    > subscript iota, should really be the corresponding uppercase letter
    > with a *subscript* iota, except in not-so-common circumstances.)

    No, this is not correct. The Unicode 4.0 and later code charts for the Greek Extended
    block definitely have this right: the correct form of e.g. U+1F08 (uppercase Alpha with
    psili breathing mark) plus U+0345 the combining subscript iota is with an adscript iota.
    [Okay, so the actual form shown in the code charts isn't very good, because the adscript
    iota is normally identical to the regular iota, but the idea is correct: it certainly
    isn't a subscript iota.]

    All of the instances of uppercase Greek character plus subscript iota which are part of
    the standard polytonic system are encoded in Unicode as precomposed diacritics. These are
    very much the 'common circumstances'. It is the not-so-common circumstances or fallback
    circumstances that I was suggesting should be handled at the glyph level. These is the
    kind of question we deal with in fonts all the time: 'How should this sequence of
    characters be rendered?' and in the case of sequences that are outside the orthographic
    norms of a writing system one has to take an intuitive approach. In this case we have a
    typographic rule for subscript iota following those uppercase letters that it normally
    follows -- i.e. that it properly takes an adscript form --, so the question is whether to
    apply this same rule, via font lookups, when it follows other uppercase Greek letters. As
    I say, this constitutes fallback circumstances, since such sequences are outside the
    bounds of the orthography. My inclination is to say, yes, this rule should apply in those
    circumstances, not least because if Unicode had not separately encoded these diacritics we
    would need to handle all complex shaping for Greek in the font lookups as we do, for
    instance, for smallcaps.

    John Hudson

    Tiro Typeworks
    Vancouver, BC
    I am not yet so lost in lexicography, as to forget
    that words are the daughters of earth, and that things
    are the sons of heaven.  - Samuel Johnson

    This archive was generated by hypermail 2.1.5 : Fri Jun 16 2006 - 22:18:02 CDT