Re: Glyphs for German quotation marks

From: Philippe Verdy (
Date: Thu Jun 08 2006 - 09:48:52 CDT

  • Next message: Adam Twardoch: "Re: Glyphs for German quotation marks"

    From: "Henrik Theiling" <>
    > However, in Courier New and Verdana, the shapes were not derived by
    > rotation, but by mirroring, which is inconsistent with Times New and
    > Arial (i.e., disregarding any definition, glyph derivation is
    > inconsistent in the realm of MS fonts). This was probably done since
    > it looks ok for English, as someone thought. English then looks like
    > this:
    > \\ //
    > --Text --

    It is Ok also for French, and perceived as equivalent to:
       66 99
          --Text --

    > Still, it is not standard compliant.

    Where have you seen that the standard said it MUST be rotated and not mirrored? The normative names DO NOT force a normative behavior. The glyphs are informative and display the recommanded 6/9-shaped forms, but the usage notes, when present, fix the normal usage. The additional resource un the standard is the mirroring property of characters, but it only specifies which character should be mirrored in RTL contexts, but it does not indicate which character should be used (mirroring is a property of the character, not of its glyph itself, because mirrored characters actually have TWO glyphs).

    There's NO such sentence in the standard that fixes compliance about the exact appearance of this character (remember that the standard does not mandate glyphs, not even according to their normative names, but just demands that the chosen glyphs cause no confusion with another character). The only caveat that exist in Microsoft fonts is for the SINGLE 6-shaped upper quote which has dfferent left or right orientation depending on fonts (and also from fallback fonts)

    > U+201c must not be a mirrored
    > shape of U+201d, but it must be rotated by 180 deg. The mirrored
    > shape on the left is called U+201F. With U+201C, it should look like
    > this:
    > // //
    > --Text --

    It would be wrong in English and French.
    More exactly, the shape would be more like:
    > _ _
    > /| /| |/ |/
    > --Text --
    when the stroke width varies (Verdana) and is not constant (Courier)

    And this would become consistant with the 9 and 6 shaped glyphs,
    > / / _ _
    > /| /| |/ |/
    > --Text -- / /
    which are just a bit curved to become like:
    > 6 6 9 9
    > --Text --

    However, typically, the slash-like quotes are almost always handwitten from top to bottom, unlike the curly quotes used in fine typography (drawn from bottom to top). To avoid the confusion, the slant of the opening quotes was mirrored and this is reflected in the slash-like glyphs for the opening quote. Their orientation is different from the 6/9-shaped curly quotes in fine typography. So the mapping of the "mirrored" or "turned" concept (valid only for the curly quotes) does not apply to the simplified slash-like quotes very commonly used for handwriting with a solid pen or roll pen, instead of a brush

    (note that with the historic plum, drawing curves is quite difficult, as it requires turning the hand to avoid breaking the extremity of the plum or scratching the paper and spreading ink drops everywhere around, so quotes would be most probably drawn more like slashes or small parenthese arcs with varying stroke width). Left-handed persons would also produce different positions for the heavy weights for the same arcs. This is due to the orientation of the plum when is hold within the hand (right-handed persons tend to write in italic, and left-handed persons tend to write in counter-italic or only with roman characters).

    This archive was generated by hypermail 2.1.5 : Thu Jun 08 2006 - 10:12:15 CDT