Re: Classification; Phoenician

From: Christopher Fynn (cfynn@gmx.net)
Date: Tue May 25 2004 - 07:50:46 CDT

  • Next message: Andrew C. West: "Re: Proposal to encode dominoes and other game symbols"

    John Hudson wrote:

    > Dean Snyder wrote:
    >
    > >>It simply doesn't make
    >
    >>> sense to me that we should do different things for Semitic than we
    >>> do for Indic.
    >>
    >
    >> Is it not a factor that the Indic "scripts" are in everyday use by
    >> living
    >> communities?
    >
    >
    > Not all of them are. It is, however, a factor that the Indic scripts
    > have varying shaping behaviour, not all of which is easily addressable
    > at the glyph level. There is a net benefit to text processing and
    > display in not unifying their encoding.
    >
    > John Hudson

    All the Indic scripts have the same basic consonant and vowel set -
    with additions and adaptations in some of the scripts where they are
    used for languages not closely related to Sanskrit.

    Where there are differences in shaping, were these differences actually
    a deciding factor for separation when these scripts were encoded in
    the UCS - or is this just an assumption? I don't recall anything about
    this in the original proposals for encoding the main Indic scripts.

    I'm *not* arguing that these scripts should have been unified - but, if
    they had been, I expect Uniscribe and other shaping engines would have
    been written a little differently and the OpenType spec for these
    scripts would have been slightly different - but it would probably work
    just as well.
    After all ISCII uses the same 8 bit code points for different Indic
    scripts and applications that work with ISCII manage to handle the basic
    shaping differences between them. Once you know which script is being
    used differences in shaping can be handled contextually - of course that
    would have necessitated encoding some control characters (or using
    markup) to indicate which particular script was being used and *that*
    may have been considered. to go beyond plain text.

    Historically the way Devanagari was written in different parts of India
    exibhits quite a range of different shaping behaviours. If you want to
    allow for this variety, assuming a particular shaping behaviour may not
    be the best model to follow. Of course modern Devanagari generally
    lacks these variations in shaping - probably because the script became
    standardised with the introduction of metal type and other printing
    technology.

    - Chris



    This archive was generated by hypermail 2.1.5 : Tue May 25 2004 - 07:53:28 CDT