Re: Why people still want to encode precomposed letters

From: Adam Twardoch (
Date: Wed Nov 26 2008 - 18:54:53 CST

  • Next message: Adam Twardoch: "Re: Why people still want to encode precomposed letters"

    verdy_p wrote:
    > I remain convinced that 'kerning' should concern all types of
    > pair-specific position adjustment of anchors,

    And now seriously. I'm with John here.

    The typesetting process can be roughly described as a mechanized process
    of spatial arrangement of pieces of ink (glyphs).

    What the pieces of ink are is very different in different technology.
    The traditional typesetting technologies followed the principle that one
    character is represented by one glyph, but more flexible technologies
    have a more loose relation between characters and glyphs. For example,
    in DecoType's ACE technologies (implemented in Winsoft Tasmeem, a
    specialized high-quality Arabic typesetting plugin for Adobe InDesign
    Middle East) the glyphs (pieces of ink) represent small portions of
    Arabic characters, and are arranged together using a sophisticated
    system of anchors.

    But calling this kerning is confusing. Most professionals working in the
    field of font technology, font development, type design and typography
    think of kerning as the traditional one-dimensional pairwise adjustment.
    A _specific_ instance of inter-glyph spatial arrangement rules.

    OpenType anchor-based glyph positioning (that can happen depending on
    chaining contexts so it is not just "pairwise" adjustment) is a more
    generalized instance of inter-glyph spatial arrangement rules that goes
    beyond kerning.

    In general, I am a friend of linguistic development and of
    terminological recycling. For example, the word "font" used to met "a
    set of metal type sorts in one face and size" and then later became a
    term for any media that carries a typeface (photographic matrices,
    digital files). So the semantic field of the word "font" has been
    expanded. But this was possible largely due to the fact that new
    technologies pushed out the old ones, so one could say that it was more
    of a semantic shift and not expansion of the term.

    With kerning, I agree that in 20 years' time, perhaps you'll be able to
    use the term "kerning" to denote any sort of inter-glyph spatial
    arrangement that happens in typesetting. But for now, this is confusing.

    The "traditional" meaning of kerning is still very alive because quite a
    few typesetting technologies still follow the old model. "Kerning" is
    The Old and "anchor-based adjustment" is The New.

    If you start using the term "kerning" to denote both The Old and The
    New, you're confusing a large part of your readers -- because those
    readers still live with the old concept of kerning in mind.

    So my recommendation is to stop denoting anchor-based adjustments as
    "kerning" or, if you really have to, to annotate your use of the term
    "kerning" with a supporting expression such as "kerning in a very broad


    Adam Twardoch
    | Language Typography Unicode Fonts OpenType
    | | |
    I hate to advocate drugs, alcohol, violence, or
    insanity to anyone, but they've always worked for me.
    (Hunter S. Thompson)

    This archive was generated by hypermail 2.1.5 : Wed Nov 26 2008 - 18:57:22 CST