From: Asmus Freytag (
Date: Mon Oct 24 2005 - 14:30:07 CST

  • Next message: David Starner: "Re: LAO LETTER FO SUNG and LAO LETTER FO TAM"

    This discussion is largely irrelevant at this stage.

    The original impetus for elevating the status of character names came
    from ISO/IEC JTC1/SC2 who wanted to use the names to identify the
    cross-mappings among SC2 standards. Names across SC2 standards were
    aligned as part of that process.

    In the mid 90's there was also an effort to design a representation for
    a character name that could be an identifier using ASN.1 (Abstract
    Syntax Notation). Whether this has lead to any further work is unknown,
    but shows that those working on the concept of character names as
    identifiers at the time were quite serious about general applicability
    of the concept.

    Around the same time WG2 bogged down in efforts or 'correcting'
    character names, led by the Scandinavian countries. It was soon realized
    that there were many names for which there was no disagreement about
    what character was being referred to, but a lot of disagreement about
    the most perfect name.

    This helped convince even people who were not interested as deeply in
    the identifier concept to see virtue in freezing the names, lest they be
    subject to the whim of shifting majorities in the committee.

    All of that is at least 10 years in the past, which means that we have
    now had a ten-year history of an unknown number of implementations that
    use names as identifiers and/or otherwise rely on stable names. From a
    practical point of view that means that we are now unable to change that
    aspect of the standard.

    The train has left the station on that one, and no amount of opining on
    whether names *should* be used as identifiers is going to matter here.

    However, their use a the (sole) non-graphical and non-numerical way to
    transmit character identity to end-users should be questioned, and
    implementations have a responsibility to understand the difference
    between 'formal' and descriptive names of characters.


    On 10/24/2005 5:53 AM, Philippe Verdy wrote:

    > From: "Richard Wordingham" <>
    >> What if header files for an application are automatically generated
    >> from UnicodeData.txt? (I for one am on the verge of using such a
    >> header file.)
    > Would you really do that so much blindly? This would mean importing
    > tend of thousands of identifiers in your program, which would cause
    > much more conflicts. If they are imported in a namespace, this will
    > clutter your program for nothing, with excessive and unneeded
    > compilation time due to the creation of management of large symbol
    > tables filled by your generated names and the true identifiers used by
    > your application or by imported system identifiers.
    > Note that Unicodeand ISO/IEC 10646 will not guarantee that Unicode
    > character names will not conflict with system identifiers or syntaxic
    > words or system API names, whatever the renaming scheme you use to
    > generate the identifiers from Unicode character names.
    > Really you will define carefully only some identifiers for characters
    > you need to reference specifically in your program, and the name will
    > you choose for those identifiers has no link to the ISO/IEC 10646
    > normative character name. You will define its value to the codepoint.
    > So you don't need those character names. Really.

    This archive was generated by hypermail 2.1.5 : Mon Oct 24 2005 - 14:31:16 CST