Re: ZWJ, ZWNJ and VS in Latin and other Greek-derived scripts

From: Asmus Freytag (
Date: Thu Jan 25 2007 - 23:00:34 CST

  • Next message: Asmus Freytag: "Re: Proposing UTF-21/24"

    On 1/25/2007 7:22 PM, John Hudson wrote:
    > Asmus Freytag wrote:
    >>> In Latin typography, ligature formation is largely a matter of
    >>> stylistic preference. Stylistic preferences do not belong in plain
    >>> text.
    >> In some styles, ligature formation is subject to additional rules
    >> that are then no longer a stylistic preference. The most widely know,
    >> if only historically relevant, of these is Fraktur (which is encoded
    >> in the Latin script). In Fraktur style, and probably dependent on the
    >> language, there are a number of ligatures that are essentially
    >> mandatory, such as 'ch' for Fraktur texts in German, and others, such
    >> as 'st' that are mandatory in some words, and prohibited in others.
    >> Once you decide to use Fraktur, ligatures become part of your
    >> orthography.
    > Right, but deciding to use fraktur is itself a stylistic preference.
    > I'm not sure that going through a text that one has decided to set in
    > fraktur -- or which might possible be displayed in Fraktur -- and
    > inserting ZWJ everywhere one wanted ligation to occure
    not necessary as the problem can be solved by having the layout rules
    supply all the mandatory ligatures and as many of the optional ones as
    needed for justification and other layout-related reasons - it's marking
    the prohibited locations that's the issue in both Fraktur-based and
    standard (!) orthography.
    > and/or ZWNJ everywhere one didn't want it is a sensible way to enable
    > the orthographic impact of this deicision.
    > The sad fact is that there is no standard layout model for fraktur,
    > nothing to say how a fraktur font should operate or how application of
    > layout features in fraktur fonts should be handled. There isn't even
    > any way to signal to an application or layout engine that a given font
    > is fraktur.
    Well put.

    By not treating Fraktur as a separately encoded script there is no
    method to use the character encoding to make that signal. (And I'm not
    saying that aspect should be different). But, as is often the case, many
    features that we assume do have a place in the character/glyph model,
    actually have no place at all, since the "glyph" part (i.e. font/layout)
    of that model is so lacking in standardization compared to the strong
    standardization on the character side of the model. (And in that
    context, the failure of something as relatively unambitious as CSS to be
    fully supported, speaks volumes).

    The situation is by no means unique to Fraktur or the Latin script, but
    I like to remind people that with all our advances we are still unable
    to simply reprint common books the way they were on the bookshelves less
    than one hundred years ago - even for a comparatively well-understood
    and widely used language such as German.

    > John Hudson

    This archive was generated by hypermail 2.1.5 : Thu Jan 25 2007 - 23:01:45 CST