At 13:52 23/10/1996 -0700, Dr. L. Rucker wrote:
>>At 12:33 23/10/1996 -0700, Rick McGowan wrote:
>>>Michael --
>>>
>>>The Tibetan half-numbers, from what I could gather speaking with the
Tibetan
>>>group who visited San Jose last year, have no "inherent" numerical value as
>>>such. They represent subtracting 0.5 (or one-half) from the number in
which
>>>they appear as the last digit; as explained in the standard. Maybe that's
>>not
>>>clear enough.
>>>
>>>For example a string like:
>>> digit-9 digit-8 halfdigit-3 (9, 8, /3)
>>>
>>>represents the number
>>> nine hundred eighty two and five tenths
>>>or
>>> nine hundred eighty two point five
>>
>>
>>Michael, this calls for 2 simultaneous things in a future version of ISO/IEC
>>14651 (when ISO/IEC 10646-1 will be published to document these characters):
>>
>>1-by default, sort these like special characters, in order of course
>>2-suggest in a non normative annex as an example documenting this, that
>> prehandling be used to replace the last digit of a series by an in-between
>> = (1/2) character following the preceding digit, as Rick explains very
well.
>>
>>Alain LaBonti
>>
>>cc UNICODE, Rick, SC22/WG20
>>
>>
>>
>Why is there need to accommodate this variant as a standard, except in terms
>of subroutines for accurate input and output uniquely for those users who
>wish to use them? I may be missing the point, but this seems to be a simple
>numerical formatting issue for a specific and limited input group. The data
>standard can remain unchanged and still allow for separate "numerical
>translation" subroutines without disturbing that standard.
We're not talking about formatting but about ordering these character
strings. It does not disturb any standard.
Alain LaBonti
This archive was generated by hypermail 2.1.2 : Tue Jul 10 2001 - 17:20:32 EDT