Re: Abuse of Unicode name list?

From: Peter Kirk (
Date: Sat Apr 09 2005 - 12:21:56 CST

  • Next message: Séamas Ó Brógáin: "Re: Abuse of Unicode name list?"

    On 06/04/2005 03:41, John Hudson wrote:

    > Peter Kirk wrote:
    >> In the light of the discussion about misleading Unicode names for
    >> Tamil etc, and the insistence of some that this is not a problem
    >> because these names are not supposed to be informative, is what MSKLC
    >> does an abuse of the Unicode name list?
    > No, it is not. MSKLC's use of Unicode names is *informative* too. ...

    Not "too", for I wrote "these names are *not* supposed to be
    informative", emphasis added. At least, this is what we have been told,
    and in fact the Unicode character names are known to be misinformative,
    to coin a word. And my concern is actually more seriously about MS
    Office, which is used by millions of naive end users, that about the
    system tool MSKLC which is for use by specialists.

    > ... It is a mnemonic aid to confirming which Unicode character you are
    > mapping to a particular key state. What would you prefer? That MS
    > invent its own character names, confusing users and generating ambiguity?

    No, I would prefer that the Unicode consortium provided a proper list of
    informative character names. And since they are unable to change the
    existing character names, they should deprecate that entire list of
    character names as unreliable and effectively replace it with a list of
    properly informative character identifiers, whose absolute stability is
    not guaranteed.

    After all, it is quite possible that a character's name may actually
    change - for example, if a country changes the official name of its
    currency unit but does not change the symbol, as Ireland did when it
    changed the name of its currency from pound to punt.

    As a first step towards such a list, Unicode should produce a formal
    database of character name corrections, listing the characters whose
    names to be used in user interfaces should be different from the formal
    character names. And Microsoft and other software companies would then
    be able to use a corrected list of names which actually provides valid
    information to its end users.

    Peter Kirk (personal) (work)

    This archive was generated by hypermail 2.1.5 : Sat Apr 09 2005 - 12:50:46 CST