Re: Gwoyeu Romatzyh marking the optional neutral tone

From: Asmus Freytag (
Date: Tue Jul 14 2009 - 01:14:48 CDT

  • Next message: Christoph Burgmer: "GR and letter case Was: Gwoyeu Romatzyh marking the optional neutral tone"

    On 7/13/2009 10:36 PM, Joe wrote:
    >>>> In that case, Unicode has just what the doctor ordered:
    >>> I beg to differ. What we need is *proportional* ideographic full stop
    >>> (neither full with nor half width, i.e. not monospaced/fixed width).
    >> The characteristic of the two ideographic full stops is that they sit in
    >> the left side of the character cell (at least that's a "typical" glyph
    >> representation for them, even if variations occur). If that one-sidedness
    >> is what you object to, then, in my view, you would indeed need a new character.
    > A./,
    > Would it be correct to say that what is being discussed here is a character the same as U+FF61 HALFWIDTH IDEOGRAPHIC FULL STOP except having the EastAsianWidth property Na(rrow), same as plain old U+002E FULL STOP and the rest of ASCII?
    > That would seem to represent Chao's presumed intention: his usage in Gwoyeu Romatzyh is, after all, specifically in the context of a romanization.
    > Jou (Pinyin-free!)
    Here's the way I understand it (and I've jumped in sideways).

    Fullwidth ideographic stop looks like this:

    | o |

    Halfwidth looks like this

    |o |

    in other words, the ball is in left field.

    if you want just a plain

    | o |

    with very little side bearings on either side, then using one of the
    existing characters for that purpose spells trouble, since your
    preference goes counter to the expectation of font suppliers, rendering
    engine implementers etc.

    IFF you were to code a new character, that one would not get Na, but N
    (neutral), which comes to the same thing as you describe. It would
    become a character that's explicitly defined NOT to be an East Asian

    Now, to my view, the circle sits lower on the baseline than you get with
    just putting the fullwidth full stop. In some software the Asian
    characters might sit on a lower baseline, but that's far from universal,
    so you are likely to see the full stop resting *on* the baseline.

    This is not the case in the samples that the first poster provided -
    here these characters are *bisected* by the base line.

    What you have is a typographically the same thing as if you took the
    U+00B0 DEGREE SIGN and moved its circle down from its superscript
    position into a subscript position.

    The two characters that come closest are U+02F3 MODIFIER LETTER LOW RING

    The latter is a combining mark (intended presumably for ideographs - and
    therefore suspect in terms of whether typical implementations would
    yield correct alignment with Latin letters). However, the placement of
    this character relative to the baseline is close to what the samples
    show - at least in some fonts.

    The former may be too low: the sample glyph in the Unicode code charts
    rests entirely below the baseline - depending on the font, even quite
    far below.

    A new character,


    would be my recommendation


    This archive was generated by hypermail 2.1.5 : Tue Jul 14 2009 - 01:17:46 CDT