SC22/WG20 N682

Dear SC22 WG20 Convenor and the ISO/IEC 14651 editors

	cc: sc22 wg20 experts

Please accept the following as an contribution to the May meeting.

This document is accepted as an official contribution of Japan 
in the meeting held on 1999-04-15 convened by Takata.

  Koreaki Fujimura, Intelligent Systems Division, Electrotechnical Laboratory,
  1-1-4, Umezono, Tsukuba, Ibaraki, 305-8568, Japan
  Tel: +81 298 54 5181, Fax: +81 298 54 5989, E-mail: fujimura@etl.go.jp

-- cut here -----

Japan's supplementary comments on FCD 14651.2

J.20) Errors in the ballot:

There were two comments with the same numbering

        J.15-21, I5) It should be explained how to deal with multiple ...

and  

        J.15-21, I6) The term "hex_symbol" does not appear in BNF.

We withdraw the latter comment.

Moreover, we withdraw the comment 

        J.15-22, I6) The sentence 
        
        All hex_symbol's are assumed to map to an integral
        ...


J.21) 4.1 canonical form:

This definition should be removed because it is not used at all.


J.22) 4.3 collation:    

The definition 

        ordering of elements

is too simple.  It should be removed or it should be changed to 

        collation: The logical ordering of strings according to defined 
        precedence rules

which has been used in POSIX and been inherited in 14652.


J.23) 4.4 collating symbol:

The definition 

        a symbol used to specify weights assigned to a character        

is not correct because one symbol specifies only one weight and weights
may be assigned to a group of characters.  It should be removed or it
should be changed to

        a symbol used to specify a weight assigned to a collating
element

        
J.24) 4.5 collating element:

The definition

        a single weight or a sequence of weights attributed to a
character 
        at a specific level of ordering 

is wrong because a single weight or a sequence of weights is assigned to
a collating element but a collating element itself is not a weight or a
sequence of weights.  It should be changed to 

        the smallest entity used to determine the logical ordering of
strings.
        A collating element shall consist of either a single character, 
        or two or more characters collating as a single entity. 

which has been developed in 14652.


J.25) 4.9 level:

The definition

        whenever used without qualification in this International
Standard, 
        level stands for the depth at which a comparison is made on two 
        character strings.

differs from the concept of the reference method which uses only one
comparison on a pair of numeric keys which are derived from some levels
of processing.  The definition should be removed.


J.26) 4.11 ordering:

The definition

        a process in which a set of strings are assigned a given order 
        relative to any other set of strings

is not understandable.  The old definition in FCD 14651.1

        a process in which a set of fields composing a record are
assigned 
        a given order relative to any other set of fields composing
other 
        records of a file

looks better.


J.27) 4.12 ordering key:

The definition

        a series of numerical values used to determine an order

differs from its usage in the reference method which uses some subkeys
but uses  only one numeric value as an ordering key.


J.28) Table syntax:

In the current table syntax, some user may use new symbols like <SABEA>
or <SCADA> which will cause an unexpected interference with the symbols
alraedy defined in the Common Template Table as 

        <SAC00>..<SD7A3> % Symbols for Hangul

Considering that FCD 14651.2 become independent of "charmap" facility,
it is better to represent hexadecimal identifiers as <_Uxxxx> or
<_Syyyy>.


J.29) Annex B.2 Example 2 - Danish delta and benchmark:

In addition to the reason described in J.17, this example is wrong
because it contains illegal symbols e.g. <//>.
This exmaple, which is written in 14652 syntax and makes sense only in a
FDCC-set with a charmap or repertoiremap, should not be included in this
standard.


J.30) 5 Symbols and abbreviations:

The contents of this clause should be moved to the next clause
"Requirements", because the contents here are not the materials used to
describe the document effectively but the materials defined as the
technical contents of this standard.


J.31) 5 Symbols and abbreviations:

The term "addenda" should be changed to "amendments".

END