Re: Questions about diacritics

From: Asmus Freytag (asmusf@ix.netcom.com)
Date: Wed Sep 15 2004 - 01:15:56 CDT

  • Next message: Asmus Freytag: "Re: TR29 Word Break awkwardness"

    At 05:21 PM 9/14/2004, Anto'nio Martins-Tuva'lkin wrote:
    >On 2004.09.14, 17:06, Jörg Knappen <knappen@uni-mainz.de> wrote:
    >
    > > My classic for this situation is the german -burg abbreviature often
    > > seen in cartography: It is -bg. with breve between b and g.
    >
    >Why not U+0062 U+035D U+0067 ? I guess that the typical presentation
    >of this convention uses a regular breve, not a "double width" one, but
    >wouldn't this be just a glyph issue?

    No - to the person encoding the text, its a text representation problem.

    The flip answer, "it's just a glyph issue" means that there is some other
    technology whose task it is to carry the intent of the user. There are many
    situations where such technology and the attendant protocols are well
    established. For example, if someone wants to change the font color of a
    word, there is style markup to do that, and some marked up (or 'rich') text
    format is the correct choice.

    For certain scripts (as well as certain notational systems) the layout
    system must implement internally some specialized rules for glyph selection
    and formatting. In the context of such scripts or systems, one can often
    rely on the existence, or anticipated existence of technology implementing
    such rules when pushing off a problem as a 'glyph issue'.

    Note that such technologies are limited to *implicit* rules and may need
    additional character codes for special characters to disambiguate some
    contexts (for example, RLM for bidi, ZWNJ for joining, SHY for hyphenation,
    etc.).

    Sometimes, glyph selection means, simply, font selection. Font selection is
    very appropriate in two situations. One, when the user wants to change the
    appearance of all instances of a given character, or even all instances of
    all characters, in a text, or some extended run of text. The other is when
    users use specialized fonts for particular symbols; either to get alternate
    representation (such as a different set of glyphs for the astrological
    symbols), or to use font technology for symbols for which Unicode does not
    provide a semantic encoding (e.g. Webdings).

    However, there are some scenarios that do not fit well with explicit font
    selection and at the same time are not currently covered by existing (or
    anticipated) layout systems, nor by existing or anticipated conventions for
    markup. The example given is a case in point.
    (neither markup, nor font switchin is currently an effective solution for
    placing a regular sized accent "between" two letters - for double wide
    accents we do have solutions in the encoding).

    In cases like that it does not help to declare that an issue is a 'glyph
    issue'. That doesn't solve the problem, but merely pushes it around. It
    would be much more helpful to at least acknowledge that neither encoding,
    nor alternate technologies can solve the problem as stated. That would
    allow us to start the search for the correct place in the character glyph
    model...

    A./



    This archive was generated by hypermail 2.1.5 : Wed Sep 15 2004 - 01:16:52 CDT